Sediment suppresses herbivory across a coral reef depth gradient

Sediments are a ubiquitous feature of all coral reefs, yet our understanding of how they affect complex ecological processes on coral reefs is limited. Sediment in algal turfs has been shown to suppress herbivory by coral reef fishes on high-sediment, low-herbivory reef flats. Here, we investigate the role of sediment in suppressing herbivory across a depth gradient (reef base, crest and flat) by observing fish feeding following benthic sediment reductions. We found that sediment suppresses herbivory across all reef zones. Even slight reductions on the reef crest, which has 35 times less sediment than the reef flat, resulted in over 1800 more herbivore bites (h⁻¹ m⁻²). The Acanthuridae (surgeonfishes) were responsible for over 80 per cent of all bites observed, and on the reef crest and flat took over 1500 more bites (h⁻¹ m⁻²) when sediment load was reduced. These findings highlight the role of natural sediment loads in shaping coral reef herbivory and suggest that changes in benthic sediment loads could directly impair reef resilience.     

Ecological Consequences of Sediment on High-Energy Coral Reefs

Sediments are widely accepted as a threat to coral reefs but our understanding of their ecological impacts is limited. Evidence has suggested that benthic sediments bound within the epilithic algal matrix (EAM) suppress reef fish herbivory, a key ecological process maintaining reef resilience. An experimental combination of caging and sediment addition treatments were used to investigate the effects of sediment pulses on herbivory and EAMs and to determine whether sediment addition could trigger a positive-feedback loop, leading to deep, sediment-rich turfs. A 1-week pulsed sediment addition resulted in rapid increases in algal turf length with effects comparable to those seen in herbivore exclusion cages. Contrary to the hypothesised positive-feedback mechanism, benthic sediment loads returned to natural levels within 3 weeks, however, the EAM turfs remained almost 60% longer for at least 3 months. While reduced herbivore density is widely understood to be a major threat to reefs, we show that acute disturbances to reef sediments elicit similar ecological responses in the EAM. With reefs increasingly threatened by both reductions in herbivore biomass and altered sediment fluxes, the development of longer turfs may become more common on coral reefs.     

Responses of algal communities to gradients in herbivore biomass and water quality in Marovo Lagoon, Solomon Islands

Settlement tiles were used to characterise and quantify coral reef associated algal communities along water quality and herbivory gradients from terrestrial influenced near shore sites to oceanic passage sites in Marovo Lagoon, the Solomon Islands. After 6 months, settlement tile communities from inshore reefs were dominated by high biomass algal turfs (filamentous algae and cyanobacteria) whereas tiles located on offshore reefs were characterised by a mixed low biomass community of calcareous crustose algae, fleshy crustose algae and bare tile. The exclusion of macrograzers, via caging of tiles, on the outer reef sites resulted in the development of an algal turf community similar to that observed on inshore reefs. Caging on the inshore reef tiles had a limited impact on community composition or biomass. Water quality and herbivorous fish biomass were quantified at each site to elucidate factors that might influence algal community structure across the lagoon. Herbivore biomass was the dominant driver of algal community structure. Algal biomass on the other hand was controlled by both herbivory and water quality (particularly dissolved nutrients). This study demonstrates that algal communities on settlement tiles are an indicator capable of integrating the impacts of water quality and herbivory over a small spatial scale (kilometres) and short temporal scale (months), where other environmental drivers (current, light, regional variability) are constant.     

Empirical relationships among resilience indicators on Micronesian reefs

A process-orientated understanding of ecosystems usually starts with an exploratory analysis of empirical relationships among potential drivers and state variables. While relationships among herbivory, algal cover, and coral recruitment, have been explored in the Caribbean, the nature of such relationships in the Pacific appears to be variable or unclear. Here, we examine potential drivers structuring the benthos and herbivorous fish assemblages of outer-shelf reefs in Micronesia (Palau, Guam and Pohnpei). Surveys were stratified by wave exposure and protection from fishing. High biomass of most herbivores was favoured by high wave exposure. High abundance of large-bodied scarids was associated with low turf abundance, high coral cover, and marine reserves. The remaining herbivores were more abundant in reefs with low coral cover, possibly because space and hence food limitation occur in high-coral-cover reefs. Rugosity had no detectable effect on herbivorous fish abundance once differences in exposure and coral cover were accounted for. At identical depths, high wave exposure was associated with greater volumes (cover × canopy height) of macroalgae and algal turfs, which most likely resulted from high primary productivity driven by flow. In exposed areas, macroalgal cover declined as the acanthurid biomass increased. The volume of algal turfs was negatively associated with coral cover and herbivore biomass. In turn, high coral cover and herbivore biomass are likely to intensify grazing. The density of juvenile corals was variable where macroalgal cover was low but was confined to lower densities where macroalgal cover was high. High coral cover and density of juvenile corals were favoured in sheltered habitats. While a weak positive relationship was found between scarid biomass and juvenile coral density, we hypothesise that high scarid densities may hinder juvenile density through increased corallivory. New hypotheses emerged that will help clarify the role of acanthurids, wave exposure, and corallivory in driving the recovery of Pacific coral communities.     

Spatial variation in the effects of grazing on epilithic algal turfs on the Great Barrier Reef, Australia

Of all benthic components on tropical reefs, algal turfs are the most widespread and the main source of primary productivity. We compared the importance of grazing by herbivores on algal turfs on two zones with marked differences in terms of benthic composition, herbivore biomass and grazing pressure, the inner flat and crest, of an inshore reef on the Great Barrier Reef, Australia. A combination of herbivore exclusion cages and transplants of coral rubble covered by algal turfs between reef zones was used to examine changes in algal turfs over a 4-day experimental period. In situ crest turfs had lower algal height, sediment loads and particulate content than reef flat turfs. Caged samples on the crest exhibited an increase in all three variables. In contrast, in situ and caged treatments on the flat presented algal turfs with similar values for the three analysed variables, with high algal height and heavy particulate and sediment loads. In the absence of cages, reef flat turfs transplanted to the crest had decreased algal height, total particulate material and particulate inorganic content, while the opposite was found in crest turf samples transplanted to the flat. Our results highlight the dynamic nature of algal turfs and the clear differences in the relative importance of herbivory in shaping turf length and sediment load between the reef crest and inner flat.     

Overgrowth and killing of corals by the brown alga Lobophora hederacea (Dictyotales,Phaeophyceae) on healthy reefs in New Caledonia: A new case of the epizoism syndrome

Coral reef degradation is often associated with regime shifts from coral‐ to macroalgal‐dominated reefs. These shifts demonstrate that under certain conditions (e.g. coral mortality, decrease in herbivory, increased nutrients supply) some macroalgae may overgrow corals. The outcome of the competition is dependent on algal aggressiveness and the coral susceptibility. In undisturbed reefs, herbivore grazing is regulating macroalgal cover, thus preventing the latter from overgrowing corals. However, some macroalgae have evolved strategies not only to outcompete corals but also to escape herbivory to some extent, allowing overgrowth of some coral species in undisturbed reefs. Epizoism represents one of those successful strategies, and has been previously documented with red algae, cyanobacteria and Lobophora variegata (Dictyotales, Phaeophyceae). Here we report a new case of epizoism leading to coral mortality, involving a recently described species of Lobophora, L. hederacea, overgrowing the coral Seriatopora caliendrum (Pocilloporidae) in undisturbed reefs in New Caledonia.     

Macroalgal Extracts Induce Bacterial Assemblage Shifts and Sublethal Tissue Stress in Caribbean Corals

Benthic macroalgae can be abundant on present-day coral reefs, especially where rates of herbivory are low and/or dissolved nutrients are high. This study investigated the impact of macroalgal extracts on both coral-associated bacterial assemblages and sublethal stress response of corals. Crude extracts and live algal thalli from common Caribbean macroalgae were applied onto the surface of Montastraea faveolata and Porites astreoides corals on reefs in both Florida and Belize. Denaturing gradient gel electrophoresis (DGGE) of 16S rRNA gene amplicons was used to examine changes in the surface mucus layer (SML) bacteria in both coral species. Some of the extracts and live algae induced detectable shifts in coral-associated bacterial assemblages. However, one aqueous extract caused the bacterial assemblages to shift to an entirely new state (Lobophora variegata), whereas other organic extracts had little to no impact (e.g. Dictyota sp.). Macroalgal extracts more frequently induced sublethal stress responses in M. faveolata than in P. astreoides corals, suggesting that cellular integrity can be negatively impacted in selected corals when comparing co-occurring species. As modern reefs experience phase-shifts to a higher abundance of macroalgae with potent chemical defenses, these macroalgae are likely impacting the composition of microbial assemblages associated with corals and affecting overall reef health in unpredicted and unprecedented ways.     

Harmful algae on tropical coral reefs: Bottom-up eutrophication and top-down herbivory

A conceptual paradigm, the “Relative Dominance Model”, provides the perspective to assess the interactive external forcing-mechanisms controlling phase shifts among the dominant benthic functional groups on tropical coral reefs [i.e., microalgal turfs and frondose macroalgae (often harmful) versus reef-building corals and calcareous coralline algae (mostly beneficial due to accretion of calcareous reef framework)]. Manipulative experiments, analyses of existing communities and bioassays tested hypotheses that the relative dominances of these functional groups are mediated by two principal controlling factors: nutrients (i.e., bottom-up control) and herbivory (i.e., top-down control). The results show that reduced nutrients alone do not preclude fleshy algal growth when herbivory is low, and high herbivory alone does not prevent fleshy algal growth when nutrients are elevated. However, reduced nutrients in combination with high herbivory virtually eliminate all forms of fleshy micro- and macro-algae. The findings reveal considerable complexity in that increases in bottom-up nutrient controls and their interactions stimulate harmful fleshy algal blooms (that can alter the abundance patterns among functional groups, even under intense herbivory); conversely, elevated nutrients inhibit the growth of ecologically beneficial reef-building corals. The results show even further complexity in that nutrients also act directly as either limiting factors (e.g., physiological stresses) or as stimulatory mechanisms (e.g., growth enhancing factors), as well as functioning indirectly by influencing competitive outcomes. Herbivory directly reduces fleshy-algal biomass, which indirectly (via competitive release) favors the expansion of grazer-resistant reef-building corals and coralline algae. Because of the sensitive nature of direct/indirect and stimulating/limiting interacting factors, coral reefs are particularly vulnerable to anthropogenic reversal effects that decrease top-down controls and, concomitantly, increase bottom-up controls, dramatically altering ecosystem resiliencies.     

Sleeping functional group drives coral-reef recovery

The world's coral reefs are in decline, with many exhibiting a phase shift from coral to macroalgal dominance . This change is often associated with habitat loss and overharvesting of herbivorous fishes, particularly parrotfishes and surgeonfishes . The challenge is to reverse this decline and enhance the resilience of coral-reef ecosystems . We demonstrate, by using a large-scale experimentally induced phase shift, that the rapid reversal from a macroalgal-dominated to a coral- and epilithic algal-dominated state was not a result of herbivory by parrotfishes or surgeonfishes. Surprisingly, phase-shift reversal was primarily driven by a single batfish species (Platax pinnatus), a fish previously regarded as an invertebrate feeder. The 43 herbivorous fishes in the local fauna played only a minor role, suggesting that biodiversity may not offer the protection we hoped for in complex ecosystems. Our findings highlight the dangers faced by coral reefs and other threatened complex ecosystems: Species or functional groups that prevent phase shifts may not be able to reverse phase shifts once they occur. Nevertheless, reversal is possible. The critical issue is to identify and protect those groups that underpin the resilience and regeneration of complex ecosystems.     

The evolution of fishes and corals on reefs: form,function and interdependence

Coral reefs are renowned for their spectacular biodiversity and the close links between fishes and corals. Despite extensive fossil records and common biogeographic histories, the evolution of these two key groups has rarely been considered together. We therefore examine recent advances in molecular phylogenetics and palaeoecology, and place the evolution of fishes and corals in a functional context. In critically reviewing the available fossil and phylogenetic evidence, we reveal a marked congruence in the evolution of the two groups. Despite one group consisting of swimming vertebrates and the other colonial symbiotic invertebrates, fishes and corals have remarkably similar evolutionary histories. In the Paleocene and Eocene [66–34 million years ago (Ma)] most modern fish and coral families were present, and both were represented by a wide range of functional morphotypes. However, there is little evidence of diversification at this time. By contrast, in the Oligocene and Miocene (34–5.3 Ma), both groups exhibited rapid lineage diversification. There is also evidence of increasing reef area, occupation of new habitats, increasing coral cover, and potentially, increasing fish abundance. Functionally, the Oligocene–Miocene is marked by the appearance of new fish and coral taxa associated with high‐turnover fast‐growth ecosystems and the colonization of reef flats. It is in this period that the functional characteristics of modern coral reefs were established. Most species, however, only arose in the last 5.3 million years (Myr; Plio–Pleistocene), with the average age of fish species being 5.3 Myr, and corals just 1.9 Myr. While these species are genetically distinct, phenotypic differences are often limited to variation in colour or minor morphological features. This suggests that the rapid increase in biodiversity during the last 5.3 Myr was not matched by changes in ecosystem function. For reef fishes, colour appears to be central to recent diversification. However, the presence of pigment patterns in the Eocene suggests that colour may not have driven recent diversification. Furthermore, the lack of functional changes in fishes or corals over the last 5 Myr raises questions over the role and importance of biodiversity in shaping the future of coral reefs.     

Macroalgal herbivory on recovering versus degrading coral reefs

Macroalgal-feeding fishes are considered to be a key functional group on coral reefs due to their role in preventing phase shifts from coral to macroalgal dominance, and potentially reversing the shift should it occur. However, assessments of macroalgal herbivory using bioassay experiments are primarily from systems with relatively high coral cover. This raises the question of whether continued functionality can be ensured in degraded systems. It is clearly important to determine whether the species that remove macroalgae on coral-dominated reefs will still be present and performing significant algal removal on macroalgal-dominated reefs. We compared the identity and effectiveness of macroalgal-feeding fishes on reefs in two conditions post-disturbance—those regenerating with high live coral cover (20–46 %) and those degrading with high macroalgal cover (57–82 %). Using filmed Sargassum bioassays, we found significantly different Sargassum biomass loss between the two conditions; mean assay weight loss due to herbivory was 27.9 ± 4.9 % on coral-dominated reefs and 2.2 ± 1.1 % on reefs with high macroalgal cover. However, once standardised for the availability of macroalgae on the reefs, the rates of removal were similar between the two reef conditions (4.8 ± 4.1 g m^?2 h^?1 on coral-dominated and 5.3 ± 2.1 g m^?2 h^?1 on macroalgal-dominated reefs). Interestingly, the Sargassum-assay consumer assemblages differed between reef conditions; nominally grazing herbivores, Siganus puelloides and Chlorurus sordidus, and the browser, Siganus sutor, dominated feeding on high coral cover reefs, whereas browsing herbivores, Naso elegans, Naso unicornis, and Leptoscarus vaigiensis, prevailed on macroalgal-dominated reefs. It appeared that macroalgal density in the surrounding habitat had a strong influence on the species driving the process of macroalgal removal. This suggests that although the function of macroalgal removal may continue, the species responsible may change with context, differing between systems that are regenerating versus degrading.     

The impact of benthic algae on the settlement of a reef-building coral

The capacity of corals to re-establish in degraded and algal-dominated habitats will depend on the effects of algae on coral settlement and growth. We tested the effect of 11 macroalgal species, of widely different functional-forms, on swimming and settlement by larvae of the coral Platygyra daedalea from the Great Barrier Reef. Algal turfs and the crustose calcareous algae groups had minor effects on coral settlement, while upright calcareous and fleshy macroalgae inhibited settlement. However, the extent of inhibition of larval settlement differed amongst upright macroalgal species, variations that were not well explained by physical differences and probably reflect chemical differences not explained by functional-form. Thus, while algal functional-form is useful in identifying general competition patterns, more detailed taxonomic and chemical approaches may be required to fully understand algal effects on corals. Different macroalgal communities on degraded reefs may have different effects on coral settlement, and hence on coral population resilience.     

Distributions of coral reef macroalgae in a back reef habitat in Moorea,French Polynesia

On tropical reefs where macroalgae are subjected to continuous herbivore pressure, spatial refuges typically are identified as large-scale, landscape interfaces that limit foraging behavior. However, algal distributions and community assemblages may also rely on the availability of smaller scale spatial refuges within the reef. The results of this study demonstrate that the patterns of macroalgal distribution across the back reef of Moorea, French Polynesia, are maintained by herbivores interacting with the small-scale structural complexities of the coral reef landscape. Although the majority of space available for colonization is composed of exposed surfaces, macroalgae rarely are found in the open. Instead, macroalgal occurrence is highest in the protected narrow crevices and hole microhabitats provided by massive Porites spp. coral heads. These distributions are determined initially by post-settlement mortality of young algal recruits in exposed habitats. Rates of consumption for two of the most common macroalgal species found in refuges across the back reef, Halimeda minima and Amansia rhodantha, indicate that algal recruits in exposed habitats are limited by herbivory. While algal abundance and community structure are highly dependent upon herbivore grazing, the availability of small-scale spatial refuges ultimately shapes the distinct community patterns and distributional boundaries of coral reef macroalgae in the back reefs of Moorea.     

Effects of physical and biological disturbances on algal turfs in Kaneohe Bay, Hawaii

Disturbance in coral reef environments commonly results in an algal community dominated by highly productive, small filamentous forms and cyanobacteria, collectively known as algal turf. Research on the types of disturbance responsible for this community structure has concentrated mainly on biological disturbance in the form of grazing, although physical and other forms of biological disturbances may be important in many coral reef areas. On the reef flat in Kaneohe Bay, Oahu, Hawaii, algal turfs grow primarily upon coral rubble that tumbles with passing swells. We manipulated the frequency of rubble tumbling in field experiments to mimic the effects of physical disturbance by abrasion and light reduction on algal biomass, canopy height, and community structure. Treatments approximated a gradient of disturbance intensities and durations that occur on the reef flat. Although sea urchins and herbivorous fishes are not widespread and abundant on the reef flat, biological disturbances to algal turf communities in the form of herbivory by small crabs and abrasion by tough macroalgae contributed significantly to the variation in algal turf biomass. Within all experiments increasing disturbance significantly reduced algal biomass and canopy heights and the community structure shifted to more disturbance-tolerant algal forms. This study shows that the chronic physical disturbances from water motion and biological disturbances other than grazing from large herbivores can control algal communities in coral reef environments.     

Do coral-reef fish faunas have a distinctive taxonomic structure?

Do the highly diverse fish faunas that associate with coral reefs have distinguishing taxonomic and ecological characteristics, as proposed by Choat and Bellwood (1991) and Bellwood (1996)? Does a 50?my old (Eocene) fossil fish fauna from Italy represent a coral-reef fish assemblage that provides unique information about the evolution of such assemblages, as claimed by Bellwood (1996)? I compared the structure of the reef fish faunas of adjacent tropical regions rich and poor in coral reefs, in both America and Polynesia, and found that they exhibit no substantive differences in relative species richness among families of typical “coral-reef” fishes. While coral-rich regions have larger reef fish faunas, a variety of factors probably contribute to such differences. Thus coral-reef fish faunas may lack a distinctive taxonomic structure. A similar comparative approach would be useful for assessing whether assemblages of fishes on coral reefs have distinctive ecological characteristics. Based on patterns of habitat use by modern tropical shorefishes, the Italian Eocene fauna includes few definite reef fishes, and may well consist primarily of non-reef fishes preserved in a non-reef habitat. Until we know more about the environment in which those fossils were preserved, that fauna can contribute little to understanding how coral reef fish assemblages have evolved.     

Interaction of herbivory and seasonality on the dynamics of Caribbean macroalgae

Many Caribbean coral reefs are undergoing a phase shift from coral to macroalgal dominance. Understanding the processes driving changes in algal abundance and community structure requires clarification of the relative effects of top-down (e.g., herbivory) and bottom-up processes (e.g., light, temperature, and nutrients). To date, a number of studies have examined the relative effects of grazing versus nutrification but interactions between herbivory and natural, seasonal fluctuations in temperature and light have not been investigated. This study considered the dynamics of three Caribbean macroalgal species [Lobophora variegata (Lamouroux), Dictyota pulchella (H?rnig and Schnetter), and Halimeda opuntia (Linnaeus)] and algal turf. A field experiment was established to measure species-specific algal dynamics (changes in abundance) over 13?months in the presence and absence of herbivory. Both herbivory and seasonal changes were important processes controlling macroalgal and turf abundance. Water temperature and light had a key role on D. pulchella; this species?? abundance significantly increased in the summer, when water temperature and light were the highest, and decreased during winter. Surprisingly, herbivory did not seem to control D. pulchella directly. However, herbivory was the most important process controlling the abundance of L. variegata, H. opuntia, and turf. The abundance of both algal species was correlated with seasonal changes in the environment, but was depleted outside cages throughout the year. The abundance of H. opuntia was positively correlated with temperature and light, but there was no statistical interaction between drivers. The statistical interaction between temperature and light was significant for the abundance of L. variegata and turf, but algal abundance declined as both factors increased. Overall, macroalgal and turf cover were mainly controlled by herbivory, while community structure (which species contributed to the overall cover) was largely influenced by seasonal changes in temperature and light.     

High herbivory despite high sediment loads on a fringing coral reef

Coral Reefs - Algal turfs are expected to increasingly dominate the benthos of coral reefs in the Anthropocene, becoming important sources of reef productivity. The sediments trapped within algal...     

Foragers versus farmers: contrasting effects of two behavioural groups of herbivores on coral reefs

Herbivorous fishes have been attributed a central role in structuring benthic communities on coral reefs. However, the relative importance of different behavioural groups of herbivores may differ and their interactions may be complex. This study focuses on an experiment that discriminates between two groups of herbivorous fish: (1) “Foragers” (relatively mobile, schooling grazers, including parrotfishes and surgeonfishes) and (2) “Farmers” (highly site-attached, territorial species, primarily damselfishes). Preliminary observations at Kimbe Bay (Papua New Guinea) showed that both groups were common, and that farmers defended areas from foragers and maintained algal communities that were distinct from the surrounding undefended substratum. An orthogonal combination of a farmer removal treatment and a forager exclusion treatment was applied to isolate their separate effects on algae and corals, and to determine whether farmer territory composition results from forager exclusion or algal cultivation. The experiment showed that foragers had quantitatively greater and qualitatively different effects on sessile benthic community structure than farmers. Where foragers were excluded, there were substantial increases in the cover and biomass of macro-algae and a decline in some corals, regardless of the presence of farmers. Where farmers were removed there was a moderate decline in the cover of some food algal species, regardless of whether foragers had access. No effect of the exclusion of foragers by farmers could be detected. Our results support prevailing views that foragers have a major impact on coral reefs and farmers cultivate selected algae, but challenge the hypothesis that damselfish influence habitat structure by moderating forager disturbance.     

Interactions between herbivorous fish guilds and their influence on algal succession on a coastal coral reef

Herbivory is an important mechanism affecting algal succession, particularly on coral reefs where the relationship between algae and corals is largely controlled by herbivores. However, different functional groups of herbivores may have contrasting effects on succession, which may explain different trajectories of coral reef recovery after disturbance. Here, the effects of different herbivore groups (roving herbivores = foragers and territorial damselfish = farmers) were isolated by a multi-factorial experiment carried out on a coastal coral reef with high macroalgal cover, high farmer densities and relatively low forager abundance. The effects of foragers and farmers were distinguished by monitoring algal succession on settlement tiles placed inside and outside exclusion cages, with orthogonal treatments established inside and outside damselfish territories (with appropriate cage controls). Within 12 months, algal assemblages on ungrazed tiles inside exclusion cages proceeded rapidly from fine filamentous turfs, to corticated algae, to tough erect (e.g. Amphiroa spp.) and foliose (e.g. Peyssonnellidae) calcified algae. Farmers had a dramatic impact on succession, essentially arresting the development of the algal community at a point where it was dominated by palatable filamentous algae of the genus Polysiphonia. Fleshy macroalgae such as Sargassum spp. were excluded from farmer territories. In contrast, foragers did not suppress fleshy macroalgae, but rather, appeared to decelerate succession and promote a relatively diverse assemblage. In contrast to forager-dominated reefs, farmer territories did not appear to function solely as forager exclusion areas or promote algal diversity as a result of intermediate grazing pressure. The relatively strong effects of farmers observed here may represent a future scenario for coral reefs that are increasingly subject to overfishing of large grazing fishes.     

Depth-dependent indicators of algal turf herbivory throughout the Main Hawaiian Islands

Herbivorous fish are key to maintaining a balance between coral and algae on reefs, where reefs with greater herbivore biomass often show lower algal cover. For reefs worldwide, algal turf cover is expanding and is increasingly used as an indicator of disturbance. Water depth affects reef fish composition; thus, it may be expected that herbivory could also differ by depth. We examined relationships between algal turf cover and biomass (g m⁻²), density (# m⁻²) and size (cm) of herbivore groups (grazers, browsers and scrapers) across shallow (< 6 m), mid (6–18 m) and deep (18–30 m) coral reefs in the Main Hawaiian Islands. We find that across all depth classes, algal turf cover decreased with increasing grazer and scraper density, with steeper relationships observed at mid and deep reefs than in shallow reefs. In contrast, algal turf cover slightly increased with increasing grazer and browser biomass at deep reefs. Considering fish size, algal turf cover increased with larger grazer and scrapers at mid and deep reefs. The results indicate that herbivorous fish density, rather than biomass, is a better indicator of reductions in algal turf cover and resulting coral-algal balance on Hawaiian reefs, where smaller fish exert greater top-down control on cover than larger fish. Despite significant differences in herbivorous fish compositions, length-frequency distributions and fishing intensities across depth, algal turf cover remains similar across depths. Increases in fishing would have a disproportionately negative impact in deep than shallow reefs due to a lower overall fish density, where grazing functions in deep reefs are maintained by significantly fewer and smaller grazers and browsers, and larger scrapers, than in shallow reefs. Developing an understanding of patterns of algal turf herbivory by depth is important to understanding the spatial scale at which herbivory and regime shifts operate.