Manta rays eat zooplankton, the small animals that drift through the water column, chiefly copepods, krill, and mysid shrimps, which they sieve from the water with plates on their gills. The daytime surface feeding that divers see is only part of the story, however. Chemical tracers in muscle tissue suggest that much of the food of oceanic mantas studied off Ecuador originates in deep water, although some may be caught near the surface at night as deep-water animals migrate upward, and that reef mantas take part of theirs from plankton living near the seafloor. Between meals, mantas visit cleaning stations, fixed spots on the reef where small fish remove their parasites, and those stations double as the social centers of manta life.
This guide is written by the scientists of the Marine Megafauna Foundation, whose research described all three manta ray species and underpins their IUCN Red List assessments.
What do manta rays eat?
Manta rays are zooplankton feeders: their diet is dominated by small crustaceans, especially copepods and krill, with mysid shrimps, larvae, and the occasional small fish. The exact menu depends on where a manta is feeding, because mantas follow whatever the local plankton is doing. At Hanifaru Bay in the Maldives, the largest known reef manta feeding aggregation, the swarms that mantas fed on were dominated by copepods, particularly the large-bodied species Undinula vulgaris, whereas the water held proportionally more gelatinous animals when mantas were present but not feeding. Off Florida’s Atlantic coast, where adult mantas now recognized as Atlantic manta rays (Mobula yarae) gather each spring, MMF’s Jessica Pate found surface plankton samples dominated by copepods, with bivalve larvae, echinoderm larvae, and arrow worms (chaetognaths) also abundant.
The most direct evidence of diet comes from stomachs, which are available only from rays killed in fisheries. In the Bohol Sea in the Philippines, where fishers have long targeted mobulid rays, a study led by MMF’s Chris Rohner examined the stomachs of four mobulid species landed between November and May. The krill Euphausia diomedeae turned up in 79% of oceanic manta stomachs, alongside copepods and a few small lanternfish (myctophids), and it was the main prey of all four species during the roughly six months a year they spend in the area. Genetic analysis of stomach contents from the same fishery, which can identify prey too digested to recognize by eye, found an unprecedented diversity of bony fishes, including 14 distinct fish sequence groups in oceanic mantas. Krill still dominated, however, and fish turned up in only one of the three sampling years, which points to occasional rather than routine fish eating.
Stable isotopes, chemical signatures in body tissue that reflect what an animal has eaten over the previous months, place reef mantas as secondary consumers, feeding on the small animals that graze on microscopic algae, and oceanic mantas at a trophic position (level in the food chain) of about 3.4, consistent with a diet built on zooplankton. For an overview of the manta species and how they differ, see the manta ray facts page and the pages on the reef manta ray, the oceanic manta ray, and the Atlantic manta ray.
How do manta rays feed?
Manta rays are ram filter feeders: they swim forward with the mouth open, unroll their cephalic lobes to funnel water inside, and strain the plankton out on their gill plates before the water leaves through the gill slits. The basic mechanism is shared by all manta and devil rays, and the anatomy behind it is covered in our anatomy chapter. What sets mantas apart is how much they vary the technique. Filtering also takes effort: drone footage analyzed by Jessica Pate and colleagues showed that feeding mantas beat their wings significantly faster than traveling mantas, even though they swam at similar speeds.
At Hanifaru Bay, researchers have classified eight distinct feeding strategies: straight, surface, chain, piggy-back, somersault, cyclone, sideways, and bottom feeding. Several are solo techniques, such as surface feeding and somersault feeding, which divers often call barrel rolling; a somersaulting manta keeps looping repeatedly within a single dense patch of plankton rather than passing through it.
Others involve groups. In chain feeding, mantas follow one another in a line; in the Maldives, hundreds of reef mantas have been reported chain feeding in a circling movement from the surface to the bottom that creates a cyclonic motion. Cyclone feeding takes this further, with many individuals stirring the water column into a vortex that concentrates the zooplankton they then feed on. Upwards of 250 mantas gather in Hanifaru Bay during peak events, and during that study, whenever plankton exceeded 200 milligrams per cubic meter, only group feeding was seen. The authors are careful about what that means, though: it is not yet clear whether the mantas are genuinely cooperating, or whether coordinated movement simply stops them colliding with each other.
When and where do manta rays feed?
Manta rays feed when plankton becomes dense enough to be worth filtering, which in practice ties feeding to particular sites, tides, weather, and seasons. Filter feeding is thought to be energetically costly, so a manta swimming open-mouthed through thin plankton could spend more energy than it gains. At Lady Elliot Island on the southern Great Barrier Reef, reef mantas began feeding only once zooplankton reached about 11.2 milligrams per cubic meter, and prey size made no difference to the decision. At Hanifaru Bay, the switch came at 53.7 milligrams per cubic meter, more than double the density estimated to meet a reef manta’s metabolic needs, which may help explain why that small bay attracts so many mantas. The Lady Elliot threshold, in contrast, sits below that estimated requirement, which suggests that mantas there must be getting additional energy from other sources, such as foraging at depth.
Tides are often what build those dense patches. At Lady Elliot Island, zooplankton biomass was highest before low tide, and about five years of sighting records showed more mantas feeding during that phase. At Hanifaru, biomass peaked just after high tide, as tidal currents drew plankton-rich water into the shallow inlet and a back-eddy trapped it there. At Egmont Atoll in the Chagos Archipelago, tagged mantas were more likely to be detected when cold-water bores pushed up the atoll slope on the flooding tide, probably carrying zooplankton up from the thermocline.
Wind matters as well. In the Maldives, tagged mantas were more likely to be detected at Hanifaru in westerly winds above 5 meters per second (11 mph), close to the new and full moon, just after high tide, conditions that may concentrate plankton through wind-driven circulation at the surface. Off northeastern New Zealand, at the southern edge of the oceanic manta’s range, daytime foraging movements were linked to northerly winds but suppressed when winds exceeded 10 meters per second (22 mph).
Feeding also varies by season, time of day, and age. Off Florida’s Atlantic coast, adult mantas aggregate each spring, typically from March through May, and the aggregation has produced the first published records of both courtship and surface prey for the species. In Indonesia, our photo-identification studies found that Manta Bay at Nusa Penida is a foraging ground used especially by juveniles, while in Komodo National Park immature mantas were most common at the northern site, Cauldron, where foraging was frequently observed (see our Indonesia manta ray program). In southern Mozambique, our acoustic receivers showed detections at a feeding site increasing at night, the reverse of the daytime pattern at nearby cleaning stations.
Do manta rays feed in the deep sea?
Partly, it seems. Chemical tracers suggest that much of the diet of oceanic mantas studied off Ecuador originates in deep (mesopelagic) water, although some of that prey may be eaten near the surface after migrating upward at night; separately, electronic tags and a research submersible have recorded mantas diving and feeding well below the surface. Reef mantas also obtain part of their food from below the surface layers where divers watch them feed. The evidence comes from three independent directions, chemistry, electronic tags, and direct observation, and each answers a slightly different question: the chemistry shows where food originated, while tags and direct observation show where mantas actually go.
The first clues came from reef mantas off eastern Australia and southern Mozambique. Our analysis of stable isotopes and fatty acids in reef manta muscle found that they do not feed solely on open-water (pelagic) plankton, and the closest match was demersal zooplankton, the animals that live near the seafloor. A companion study found unusually high levels of omega-6 fatty acids in both reef mantas and whale sharks, even though the surface crustaceans both species are seen eating are generally dominated by omega-3s, which raised new questions about the origin of their main food.
Oceanic mantas showed the pattern even more strongly. Using biopsies from 75 mantas at two aggregation sites off mainland Ecuador, our stable isotope work estimated that surface zooplankton contributed on average 27% of the diet and deep-water (mesopelagic) sources 73%; even the most conservative model put the surface share at only 43%. A follow-up study of the same population found a fatty acid profile markedly different from surface zooplankton and rich in compounds common in deep-sea organisms. That study also flagged an important caveat: many deep-water animals migrate toward the surface at night, so the chemistry cannot say at what depth the prey was actually eaten. A pilot study in the same population compared skin mucus, which may record more recent meals, with muscle and found a broader possible range of prey in mucus but no significant difference in the expected prey profile; how quickly mucus records a change in diet has yet to be measured, so different tissues may eventually reveal diet over different timescales.
Behavior helps fill that gap. Off Socorro Island in Mexico, a research submersible filmed an oceanic manta somersaulting through a thick zooplankton layer at 130 to 140 meters (430 to 460 feet), with mysid shrimps, krill, and copepods identified in the footage; tags on the same population showed mantas spending much of their time around the top of the thermocline, where plankton concentrates. Off northern Peru, tagged oceanic mantas dived deeper at night than by day, apparently to feed on vertically migrating zooplankton, returning to warm surface water between dives. Reef mantas in New Caledonia made their deepest dives, to 672 meters (2,200 feet), mostly at night, a behavior the authors suggest gives them access to food that the surface waters cannot supply. Off eastern Australia, satellite-tagged reef mantas spent significantly more time than expected in an offshore eddy that a behavioral model identified as an important foraging ground.
All of this fits the daily rhythm described in the IUCN Red List assessments: mantas often move inshore during the day to clean and socialize, then move offshore at night to feed. It also means that the surface feeding events tourists see offer only a partial view of what mantas eat, and how much any one feeding site contributes to its overall energy budget remains an open question. Our movements chapter covers diving depths and tracking in more detail.
Do manta rays and devil rays eat the same food?
Largely, yes: where several mobulid species share a feeding ground, their diets overlap heavily. Stable isotope data from manta and devil rays landed in Peru, Sri Lanka, and the Philippines, in a study co-authored by MMF’s Chris Rohner, showed dietary overlap between species sharing a site at every location studied, and the overlap appeared greatest where food was scarcest, the opposite of what ecological theory would predict. The Bohol Sea stomachs told the same story: one krill species was found in 81 of 89 stomachs (91%) across all four species, with only small differences, such as lanternfish in oceanic mantas and squid and fish in sicklefin devil rays.
This matters for more than ecology. Species that feed on the same prey, in the same places, at the same time, end up in the same nets, which may explain the multi-species mobulid bycatch recorded in fisheries around the world. It also means that identifying shared foraging grounds that overlap with fishing activity could help managers protect several threatened species at once.
Do manta rays eat plastic?
Yes. Manta rays cannot separate plastic from plankton, so where their feeding grounds are polluted, they swallow it. A review led by MMF’s Elitza Germanov set out why filter feeders are particularly exposed: they must process hundreds to thousands of cubic meters of water a day, and can take in microplastics directly from the water or indirectly through contaminated prey, along with the toxic chemicals that plastics carry.
Our field study in Indonesia then measured that exposure. Sampling manta feeding grounds at Nusa Penida and Komodo between 2016 and 2018, we found plastic abundance up to about 44 times higher in the wet season than the dry season. Combining those concentrations with estimated filtration rates, a feeding manta at Nusa Penida could swallow around 63 pieces of plastic an hour in the wet season, compared with about four an hour in the dry season. Of 22 samples of manta feces and vomit collected by local dive professionals, two contained plastic, confirming that ingestion happens and that mantas cannot selectively exclude microplastics from their food. A follow-up study led by MMF’s Janis Argeswara found that 99% of the small plastics in a critical Nusa Penida feeding ground were polyethylene or polypropylene, mostly broken down from larger items; these are relatively low-toxicity plastics in themselves, but their component chemicals, the pollutants they pick up, and the microbes growing on them remain a concern.
Remoteness offers no escape. In the Chagos Archipelago, a large and remote marine protected area, surface waters in reef manta feeding areas held an average of 1.1 microparticles per cubic meter, and Egmont Atoll, recognized as an Important Shark and Ray Area for reef mantas, was the most contaminated at 1.6. What plastic does to a manta has not yet been measured; the possible consequences include exposure to toxic plastic additives and the pollutants that stick to plastic surfaces. Because most of this plastic originates on land, so do the solutions: our Microplastics & Megafauna project has surveyed communities near Indonesian manta feeding grounds about their waste habits and shares its findings to inform local waste management policy. The wider threat picture is covered in our threats chapter.
What is a manta ray cleaning station?
A cleaning station is a fixed spot on a reef where small cleaner fish remove parasites and dead skin from visiting mantas, which return to the same stations again and again. Cleaner fish, commonly wrasses, remove ectoparasites, dead or injured skin, and mucus. That service matters, because cleaning is considered vital for maintaining health and for wound healing. At Lady Elliot Island, the distribution of one cleaner, the bluestreak cleaner wrasse (Labroides dimidiatus), predicted where tagged mantas spent their time, along with hard coral and prominent reef structures.
Cleaning takes up a surprising share of a manta’s day. In southern Mozambique, our acoustic tags showed that 89.2% of detections at cleaning stations happened in daylight, with visits averaging about 25 minutes and some lasting up to 8.2 hours; the daytime bias is likely to reflect the cleaner fish, which are typically active only during the day. At Lady Elliot Island, 81% of manta encounters within the study area involved cleaning, far more than feeding or courtship, and the authors suggest that mantas may carry a long-term memory of where good cleaning is to be found. Animal-borne video cameras on reef mantas in the Maldives recorded cleaning as the second most common daytime behavior, after cruising.
Individual stations can draw large numbers of mantas. At Red Sands, a reef cleaning station in Závora Bay, Mozambique, a study led by Michelle Carpenter and co-authored by MMF scientists identified 583 individual reef mantas between 2010 and 2021, more than half of them seen again in later years (see our Mozambique manta ray research). Devil rays use the same service: off South Africa, a study from the same lead author documented shortfin devil rays being cleaned by bluestreak cleaner wrasse at a newly discovered cleaning area, in groups of more than 150.
Not every small fish traveling with a manta is a cleaner, however. In the Maldives, reef mantas at cleaning stations and near-term pregnant females were the most likely to carry adult sharksucker remoras, and nine years of freediver surveys off south Florida found that remoras were the most frequent of four groups of hitchhiking fishes, riding mostly on the underside toward the right rear of the ray.
How do manta rays balance feeding and cleaning?
Mantas generally trade one activity off against the other, cleaning when conditions are poor for feeding, but where food and cleaners sit close together they move back and forth between the two. At a commonly fished seamount near Ticao Island in the Philippines, cameras on a cluster of cleaning stations showed mantas visiting when conditions were poor for foraging but suitable for cleaning, and staying away when plankton aggregated, with sea state, moonlight, and current speed among the predictors. In the Maldives, detections at a cleaning station increased in calmer winds below 5 meters per second (11 mph), whereas feeding at nearby Hanifaru was associated with stronger winds, which suggests that mantas use cleaning stations when conditions are not suitable for feeding.
That trade-off is not absolute. Around D’Arros Island in the Seychelles, 11 identified reef mantas made 15 feeding–cleaning cycles over four survey days, each lasting between 11 and 102 minutes, feeding at the surface as close as about 20 meters (66 feet) from the stations, even when plankton was dense. The authors argue that where cleaning stations sit right next to productive feeding areas, mantas can switch activities without losing much energy. That is consistent with the hypothesis that mantas prefer cleaning sites close to productive foraging regions.
Some sites specialize instead. At Nusa Penida, two shallow bays about 12 kilometers (7.5 miles) apart do different jobs: Manta Bay is mainly a feeding ground, while Manta Point is used mainly for cleaning and courtship, with 44% of individuals moving between them. For managers, the lesson is that feeding sites and cleaning stations function as a connected network of key habitats, and researchers in the Maldives recommend protecting the whole network, including the corridors that link its sites.
Why are cleaning stations so important to manta rays?
Because they are where mantas meet: cleaning stations are the main venues for courtship and social life, not just for grooming. Over 14 years in the Maldives, researchers recorded 229 courtship events, and 90% of them took place at cleaning sites. At Nusa Penida’s Manta Point, mature females made up 26% of the community but were the most frequently sighted group, accounting for 48% of sightings, and courtship coincided with the annual peak in sightings in May. The courtship sequence itself is covered in our reproduction chapter.
Cleaning stations also structure manta society. Acoustic tracking in Raja Ampat, Indonesia, in a study co-authored by MMF’s Stephanie Venables, found that reef mantas formed social communities tied to particular sites, which remained stable over several weeks to months, and that the most site-attached individuals had the strongest social bonds. At cleaning stations in the same region, mantas curled and uncurled their cephalic lobes in ways that changed when they approached other mantas, divers, or cleaner fish; gestural communication is the most likely explanation, although it has not been confirmed. Animal-borne cameras recorded reef mantas spending an average of 43% of their recorded time interacting with other mantas, compared with 8% for oceanic mantas, although the authors caution against generalizing beyond the two populations studied.
The same predictability that makes cleaning stations valuable to mantas makes them valuable to tourism. Divers who approach too closely can cause mantas to leave a cleaning station, and poor buoyancy control damages the coral around it, which is why codes of conduct for manta tourism matter. Our tourism chapter covers how to visit a station without disrupting it.
Do manta rays eat fish, or bite people?
Manta rays do not hunt fish and do not bite people; they are filter feeders, and their small teeth play no known part in feeding. Small fish do occasionally turn up in their stomachs. Stomach analyses show that mobulids feed on zooplankton and small fishes, and the lanternfish found in Bohol Sea oceanic mantas may represent opportunistic extra energy when dense krill patches are not available, rather than active hunting.
A manta’s teeth sit in a band on the lower jaw, and our Mozambique research notes that mantas may have retained them solely for mating, when a male grips the female’s pectoral fin. Reef mantas have no tail spine at all, and the oceanic manta’s spine is embedded in a calcified knob at the base of the tail, as set out in the 2009 redescription of the genus by MMF co-founder Dr. Andrea Marshall. Mantas are not dangerous to swimmers, and our anatomy chapter covers their teeth and tails in more detail; if anything, the risk runs the other way.
Why it matters for conservation
Feeding grounds and cleaning stations are the predictable places where mantas concentrate, which makes them both the best places to protect mantas and the places where mantas are most exposed. The Red List assessment for the oceanic manta notes that fisheries coincide with high-productivity areas where mantas are likely to aggregate to feed, and the Bohol Sea, where mobulids are caught while feeding on the same krill swarms, is exactly such a place.
The opportunity is to match protection to how mantas actually use these sites. In Mozambique, our acoustic tracking showed that marine protected areas covered only about 24% of the area reef mantas were using, and we recommended explicit protection of critical habitats at Praia do Tofo and Závora; Mozambique has since banned the capture of manta and devil rays under its 2020 marine fisheries regulations, in force from January 2021. In the Maldives, motorized boats and scuba diving are prohibited in Hanifaru Bay because of the numbers of mantas and other megafauna using it, a clear example of site management built around a feeding aggregation. Plastic pollution, in contrast, has to be tackled at its source on land, through the kind of community engagement described above.
Feeding sites and cleaning stations are also where much of the long-term monitoring happens. The photo-identification catalogs behind our population studies were built at sites such as Red Sands in Závora Bay and Manta Bay and Manta Point at Nusa Penida, so keeping those sites healthy also keeps the record of how manta populations are faring. You can adopt a manta ray to support new research on manta rays.