Skip to content
NatureHQ

species

Orca

Orcinus orca Linnaeus, 1758

Orcas are the largest dolphins, living in stable matrilineal groups with distinct vocal traditions, and are one of very few animals in which females live for decades after they stop breeding.

The orca is a dolphin, not a whale in the everyday sense, and "killer whale" comes from an old name meaning whale-killer. What makes it worth a NatureHQ record is not size but social structure. Resident orcas live in matrilineal groups that stay together for life, each with its own stable repertoire of calls that has persisted for decades. Females stop reproducing around forty and can live into their eighties — a pattern found in almost no other animal — and those older females lead group movement most often in the years when salmon are hardest to find. Different ecotypes hunt entirely differently, and those differences are learned rather than inherited, which makes orca one of the strongest cases for culture outside primates.

Developed record · 82% complete · reviewed 2026-08-09

What this page covers

One recognised species, Orcinus orca — but divided into ecotypes that differ so much in diet, society and behaviour that several may warrant species status. Nearly all the cognition research covers North Pacific residents.

Often confused with: Pseudorca crassidens (false killer whale — a different genus entirely); Globicephala (pilot whales — also large black delphinids)

Quick facts

Species
Orcinus orca — the largest member of the dolphin family
Female lifespan
Commonly into the 80s; reproduction usually ceases around 40
Vocal traditions
Each pod has its own stable call repertoire, transmitted maternally
Ecotypes
Residents, transients and offshores differ in diet, society and voice

Almost every general statement about orcas is really a statement about one ecotype. Resident orcas eat fish, live in large stable matrilines and are extremely vocal. Transients — also called Bigg’s killer whales — hunt marine mammals in small quiet groups, because seals can hear. Offshores are poorly known and appear to eat sharks. These groups do not interbreed and barely interact.

This matters for reading research the same way the crow-and-raven distinction does. A finding about resident vocal dialects says nothing about transients, and a documentary sequence of orcas beaching themselves to take seals shows a hunting tradition confined to a few populations off Patagonia and the Crozet Islands.

Related

  • Crows

    The same problem: one popular name covering several different animals

Words used here
Ecotype
A population within a species that differs consistently in diet, behaviour and often appearance, without being a separate species — at least not yet.

Killer whale females stop breeding decades before they die — and lead the group afterwards

Well supported

Good evidence backs this, though some details remain open.

Female resident killer whales cease reproduction around age 40 but may live into their 80s. Post-reproductive females lead group movement disproportionately, most strongly when salmon are scarce, and simultaneous breeding by mothers and daughters sharply reduces the older female’s calf survival.

Who this applies to
resident killer whales in long-term North Pacific study populationsDo not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Orcinus orca
Why we rate it this way, and what the caveats are
Well supportedHigh confidence

Two independent analyses of the same multi-decade individual-level record, testing different predictions — leadership under scarcity, and the survival cost of overlapping generations — and both were borne out.

How far it can be extended

Explicitly not generalised. Menopause is known in only a handful of toothed whales and humans, and the demographic conditions that explain it here are specific to this social structure.

Caveats

  • Leadership is inferred from spatial position, not from an observed decision.
  • Observational demography — no manipulation is possible or ethical.
  • One prey system; killer whales elsewhere eat very differently.

Still unanswered

  • How is ecological knowledge actually transferred from an old female to the group?
  • Why has menopause evolved in so few species?

Last reviewed 2026-08-09

The evidence (2 studies)

In resident orcas, neither sons nor daughters leave their mother’s group. That is unusual: in most mammals one sex disperses. It means a female’s relatedness to the animals around her rises steadily as she ages, and that single demographic fact turns out to explain a great deal — including why she stops breeding while her daughters are starting.

Words used here
Matriline
A female and all her descendants. In resident orcas this is the basic unit of society, and it does not break up.

Every killer whale pod has its own set of calls, passed down the mother’s line

Well supported

Good evidence backs this, though some details remain open.

Resident killer whale (Orcinus orca) pods in the North Pacific use stable, group-specific repertoires of discrete calls that persist across decades and are shared in proportion to recent maternal ancestry.

Who this applies to
resident killer whales off British ColumbiaDo not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Orcinus orca
Why we rate it this way, and what the caveats are
Well supportedHigh confidence

Repertoires recorded over roughly fifteen years across sixteen pods, with sharing tracking maternal descent. Stability across decades is difficult to explain by anything other than social transmission.

How far it can be extended

Explicitly not generalised. Resident, transient and offshore killer whales differ in diet, society and vocal behaviour, and this work covers residents only.

Caveats

  • These are called dialects by analogy. The calls are not known to carry meaning the way words do.
  • What the calls are *for* is largely unresolved.
  • Residents only — transient killer whales are far quieter, because their prey can hear them.

Still unanswered

  • Do discrete calls encode anything beyond group identity?
  • How quickly does a repertoire change when pods merge or split?

Last reviewed 2026-08-09

The evidence (1 study)

It is worth being careful with the word "dialect". What has been shown is that each pod uses a distinctive and remarkably stable set of calls, and that pods sharing recent maternal ancestry share more of them. What has not been shown is that the calls mean anything in the way words do. The analogy is to accent rather than to vocabulary.

Compare across species

  • Elephants

    Individually specific calls that function something like names

What a killer whale actually eats

Specific enough to change a conservation conclusion.

Fish-eating killer whales in the north-east Pacific eat mostly one salmon species

Well supported

Good evidence backs this, though some details remain open.

Prey remains collected at kill sites of resident killer whales in British Columbia over more than two decades were overwhelmingly Chinook salmon, despite Chinook being far less abundant than sockeye or pink salmon in the same waters — indicating active selection rather than opportunistic feeding.

Who this applies to
fish-eating resident killer whales in British Columbia and adjacent watersDo not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Orcinus orca
Why we rate it this way, and what the caveats are
Well supportedHigh confidence

Twenty-two years of prey-fragment collection at observed kills, with the selection inferred from the mismatch between diet and availability rather than from diet alone.

How far it can be extended

Applies to one ecotype in one region. Mammal-eating transients in the same water eat no fish at all, which is precisely the point.

Caveats

  • Surface fragment sampling under-represents small prey that leaves nothing to collect.
  • One region; other fish-eating populations have different specialisations.
  • Diet composition may shift as Chinook runs decline.

Still unanswered

  • Can southern residents switch prey as Chinook decline, or is the specialisation fixed by tradition?
  • How is a preference for one salmon species transmitted between generations?

Last reviewed 2026-08-09

The evidence (2 studies)

Twenty-two years of collecting scales and fragments from the water where residents had just killed something produced an unexpectedly narrow answer: overwhelmingly Chinook salmon, the largest and fattiest species and far from the most abundant. Sockeye and pink salmon, many times more numerous in the same water, were taken rarely.

"Orcas eat fish" and "these orcas depend on one declining salmon species" support entirely different conservation conclusions, and only the second is true of southern residents. It is also the plainest illustration of why the ecotypes should not be discussed as one animal: mammal-eating transients share the same water and eat no fish at all.

Dialects that drift, and what that proves

A repertoire that changed over forty years of tape.

Killer whale calls change slowly over decades — and change together within a family

Well supported

Good evidence backs this, though some details remain open.

Comparison of archived recordings spanning more than a decade shows the structure of shared calls in resident killer whale matrilines changing gradually over time. Changes are shared within a matriline rather than arising independently in individuals, and related matrilines diverge acoustically as they change.

Who this applies to
resident (fish-eating) killer whale matrilines in the north-east PacificDo not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Orcinus orca
Why we rate it this way, and what the caveats are
Well supportedModerate confidence

Long recording archives are the only way to observe drift, and the pattern — shared within a group, divergent between related groups — is the signature of learning rather than inheritance. Confidence is moderate because two matrilines were analysed and automated call classification involves methodological choices that affect the result.

How far it can be extended

Documented in resident ecotypes in one region. Transient and offshore ecotypes have different vocal behaviour and are far less studied.

Caveats

  • Two matrilines, with uneven recording effort across years.
  • Cultural drift is inferred from the pattern of change rather than observed being transmitted.
  • Applies to resident ecotypes; other ecotypes are much less studied.

Still unanswered

  • Do calves learn calls from the mother specifically, or from the matriline as a whole?
  • How much acoustic divergence is needed before two groups stop associating?

Last reviewed 2026-08-10

The evidence (2 studies)

A group-specific call repertoire could in principle be inherited. What rules that out is change: comparing archived recordings of the same matrilines across more than a decade shows call structure shifting gradually, with the shift shared inside a matriline rather than appearing independently in individuals — and related matrilines drifting apart from one another as they change.

That pattern is the signature of learning. Genes do not drift in unison within a social group over a decade and then diverge between related groups; a learned tradition does exactly that. It is also why acoustic similarity between pods can be read as a rough measure of how recently they shared a matriline.

The North Pacific recording archives that made this possible span decades. Some questions in animal behaviour can only be answered by somebody having pressed record for forty years.

Words used here
Cultural drift
Gradual change in a socially transmitted tradition, analogous to genetic drift but travelling through learning rather than inheritance.

Killer whale ecotypes are genetically separating, and the boundaries follow what they eat

Contested

Researchers actively disagree, and the disagreement is substantive.

Genome sequencing across five killer whale ecotypes shows recent divergence with little gene flow despite overlapping ranges, and signals of selection in genes associated with diet and digestion. The genetic boundaries correspond to differences in learned foraging tradition rather than to geography.

Who this applies to
five killer whale ecotypes across the Pacific and Antarctic
Studied in
Orcinus orca
Why we rate it this way, and what the caveats are
ContestedModerate confidence

The genomic divergence and its correspondence with ecotype boundaries are measured. That culture caused the divergence, rather than accompanying it, is an inference the data cannot settle — and the authors say so.

Caveats

  • Fifty individuals across five ecotypes is thin for population genomics.
  • Sampling is uneven, with North Pacific residents best represented.
  • Ecotypes are not formally recognised as separate species, and whether they should be is unsettled.

Where researchers disagree

  • Whether learned tradition drove the genetic divergence, or arose alongside it, cannot be resolved from genomes. The correlation is compatible with either direction and with a third factor driving both.

Still unanswered

  • Should the ecotypes be recognised as separate species?
  • Can the causal direction between culture and genetic divergence be established at all?

Last reviewed 2026-08-10

The evidence (3 studies)

Resident and transient killer whales share the same water off British Columbia. They are the same species, physically capable of interbreeding, and they do not. They eat different food, hunt differently, sound different and do not associate. Genome sequencing across five ecotypes finds correspondingly little gene flow, recent divergence and selection signals in genes tied to diet.

The strong claim — that culture is driving the genetic separation — is an inference the genomes cannot settle on their own, and the authors are explicit about it. The correlation is compatible with culture driving divergence, with divergence enabling cultural difference, or with something else driving both. What is not in doubt is that the boundary follows behaviour rather than geography.

This is why NatureHQ scopes almost every orca claim to an ecotype. "Orcas eat seals" and "orcas eat salmon" are both true of some populations and false of others, and a page that says "orcas do X" is nearly always describing one group.

Words used here
Ecotype
A population distinguished by diet, behaviour and vocal repertoire rather than by formal taxonomy. Killer whales have several, overlapping in range.
Gene flow
The movement of genes between populations through interbreeding. Its absence between overlapping ecotypes is the striking result here.

Killer whales sleep the way other cetaceans do — one cerebral hemisphere at a time, with the opposite eye closed, so that breathing and swimming continue throughout. Resting groups surface in synchrony and move slowly in tight formation, which is what a sleeping pod looks like from a boat.

Newborn calves and their mothers show almost no conventional sleep at all for the first weeks of life. An animal that swims continuously from birth, apparently without the rest every other mammal needs, remains one of the genuinely unexplained findings in cetacean physiology.

Related

  • Dolphins

    The same sleep architecture, better studied

  1. 1970

    First observation

    Individual photo-identification begins

    Michael Bigg established that killer whales could be identified individually from saddle patches and dorsal fin shape. Before this, populations were estimated by guesswork and assumed far larger than they were.

  2. 1991

    Landmark experiment

    Stable pod-specific dialects described

    Systematic recording established that each matriline has its own repertoire of discrete calls, shared within the group and differing between groups.

    Vocal traditions among resident killer whales (Orcinus orca) in coastal waters of British Columbia

  3. 2000

    Reinterpretation

    Dialects shown to change over time

    Archived recordings spanning more than a decade showed call structure drifting gradually, shared within matrilines and diverging between related ones — the signature of learning rather than inheritance.

    Changes how the 1991 result reads

    A stable group-specific repertoire is compatible with inheritance: related animals could simply sound alike. Change is not. Watching a dialect drift within a matriline, and two related matrilines drift apart, removes the genetic explanation.

    Dialect change in resident killer whales: implications for vocal learning and cultural transmission

  4. 2006

    Modern discovery

    Prey specialisation quantified

    Twenty-two years of prey fragments collected at kill sites showed resident killer whales feeding overwhelmingly on Chinook salmon despite far more abundant alternatives.

    Selective foraging by fish-eating killer whales Orcinus orca in British Columbia

  5. 2012

    Landmark experiment

    Culture proposed as a barrier to gene flow

    Synthesis of behavioural, acoustic and genetic evidence argued that learned traditions were themselves keeping overlapping ecotypes reproductively separate.

    Cultural traditions and the evolution of reproductive isolation: ecological speciation in killer whales?

  6. 2016

    Challenge

    Genomes show the divergence is real, and its cause still unsettled

    Fifty genomes across five ecotypes confirmed recent divergence with little gene flow and selection on diet-associated genes — while leaving the direction of causation between culture and genetics unresolved.

    Changes how the 2012 result reads

    The genomic data confirmed the pattern and refused to confirm the mechanism. Culture and genetic divergence have moved together, and observational and genomic data alike are compatible with either driving the other — which is why NatureHQ records the culture-as-barrier claim as contested rather than established.

    Genome-culture coevolution promotes rapid divergence of killer whale ecotypes

There is no confirmed record of a wild orca killing a person. There are several records of captive orcas killing their trainers. That contrast is the single most useful fact about orcas and people, and it is rarely stated plainly: the danger is not in the animal but in the conditions.

Animal welfare

Watching orcas responsibly

Vessel noise interferes with the echolocation resident orcas use to find salmon, and disturbance has been implicated in the decline of some populations. Where approach distances are set in law they exist for that reason rather than for safety. Watching from land costs the animals nothing.

Where this applies: Approach distances differ by country and by population; several North Pacific populations have specific legal protections.

The 2023 reports of orcas striking boat rudders off Iberia were widely reported as revenge or as a coordinated campaign. Neither framing is supported. The behaviour is confined to a small number of identified individuals, spread through that group, and researchers studying it are explicit that its cause is unknown — a fad, play, or a response to a bad experience are all live possibilities.

Killer whale groups hunt in completely different ways, and the differences are taught

Well supported

Good evidence backs this, though some details remain open.

Killer whale ecotypes that overlap in range differ systematically in diet, hunting technique and vocal repertoire, do not interbreed, and show corresponding genetic divergence — a pattern consistent with learned traditions acting as a barrier to gene flow.

Who this applies to
killer whale ecotypes, principally in the North Pacific and North Atlantic
Studied in
Orcinus orca
Why we rate it this way, and what the caveats are
Well supportedModerate confidence

The behavioural and genetic divergence is well documented and the range overlap makes geography an implausible explanation. Confidence is held at moderate because the direction of causation between culture and genetic separation cannot be settled observationally.

Caveats

  • Whether culture drove the genetic divergence or accompanied it is not resolved.
  • Ecotype sampling is very uneven; North Pacific residents dominate the literature.
  • A hunting technique confined to a few populations is often reported as something orcas do.

Still unanswered

  • Should the ecotypes be recognised as separate species?
  • How is a hunting technique actually transmitted to a calf?

Last reviewed 2026-08-09

The evidence (3 studies)

What an orca eats depends almost entirely on which group it was born into. Resident orcas in the North Pacific eat fish, and some populations specialise so narrowly that a decline in one salmon species threatens them directly. Transients eat marine mammals. Populations off Norway herd herring into a ball and stun them with tail slaps; populations off Patagonia and the Crozet Islands deliberately strand themselves on beaches to take seal pups, then work their way back into the water. Orcas off South Africa have been documented opening white sharks and eating the liver, leaving the rest.

None of these techniques is available to an orca that was not taught it. Beach-hunting takes years to learn and is genuinely dangerous — a mistimed attempt strands the animal — and adults have been observed pushing calves through practice runs on gently shelving beaches. This is the strongest reason the ecotypes are treated as cultural rather than merely regional.

Two cautions belong with this material. A hunting technique confined to a few populations is routinely reported as something "orcas do", which is how a Patagonian tradition becomes a fact about the species. And attacks on large whales, though real, are rare and hard-won: sperm whales meet them with an organised defensive formation, and most documented encounters do not end in a kill.

Compare across species

Words used here
Wave washing
Coordinated swimming that generates a wave to sweep a seal off an ice floe. Documented in a small number of Antarctic groups.

An orca is the largest dolphin rather than a whale in the everyday sense, and much of the body follows from being a large air-breathing predator in cold water: a thick blubber layer, a countercurrent circulation that keeps heat in the core, and lungs that exchange a far larger fraction of their volume per breath than ours do. Like all cetaceans it breathes voluntarily through a blowhole, and would drown if it lost consciousness entirely.

Hunting is where the ecotypes separate most sharply, and the techniques are learned within a matriline rather than being properties of the species. Some populations herd herring into a ball and stun them with tail slaps. Some wash seals off ice floes with a coordinated wave. Some strand themselves deliberately on a beach to take a sea lion and wriggle back — a technique taught to calves over years, in shallow water, by adults who push them back out.

  • Fish-eating populations hunt cooperatively and call constantly; mammal-eating populations hunt quietly, because their prey can hear them.
  • Some populations take the liver of large sharks and leave the rest — a fat-rich organ, and a precise piece of butchery.
  • Prey ranges from herring to blue whales depending on population. No orca population eats everything.
  • There is no reliable record of a wild orca killing a person, which is remarkable for an animal of this size and capability.

The collapsed dorsal fin of captive males is worth a note because it is so often misread. It is common in captivity and rare in the wild, and the leading explanations are the fin lacking the support that fast swimming through deep water provides, together with time spent at the surface. It is a consequence of the conditions rather than a normal feature of the animal.

The white patch behind an orca’s eye is not the eye. The actual eye sits just in front of it and slightly below, which is one reason the patch is thought to function in signalling or in misdirecting prey.

  • Length: males to about 8–9 metres, females to about 7–8. The dorsal fin of an adult male can approach 2 metres and is the easiest way to sex an animal at a distance.
  • Weight: up to around 6 tonnes for a large male.
  • Routine dives: three to five minutes, with most foraging well inside ten.
  • Maximum recorded breath-hold: around fifteen minutes, and long dives are the exception rather than the rule.
  • Depth: usually within the top 100 metres, though dives beyond 250 have been recorded in some populations.

The breath-holding figure is often inflated, and the reason it stays modest is that orcas mostly hunt near the surface, where their prey is. Capability and habit are different questions: an orca could stay down considerably longer than it usually does, and rarely has any reason to.

Words used here
Ecotype
A population with its own diet, calls and behaviour, not interbreeding with neighbours despite an overlapping range.
Countercurrent exchange
Warm outgoing blood running alongside cold returning blood, so heat is recycled rather than lost at the extremities.

A seventeen-month pregnancy, and a life measured in one population

Females outlive males by twenty years, and the famous numbers describe a single studied group.

Female orcas outlive males by decades, and the famous numbers describe one population

Established

Specialists would state this without hedging. Multiple independent lines of evidence agree.

In the northern resident population of British Columbia, decades of photo-identification give mean life expectancy for females surviving to adulthood of around 50 years with maxima in the eighties, against around 30 years for males with maxima near 60. Females first calve at about 14 years, produce a calf roughly every five years, and cease reproduction in their late thirties or early forties while frequently living decades beyond it. Gestation is approximately 17 months.

Who this applies to
the northern resident population of British ColumbiaDo not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Orcinus orca

You may have heard

Orcas live to 100

A very old female in a well-studied population may approach it, and the average is nothing like it — and males average around thirty. Quoting one figure for a species whose two sexes differ by twenty years, across populations that differ again, is how a real measurement becomes a wrong fact.

Why we rate it this way, and what the caveats are
EstablishedHigh confidence

Every individual is identifiable from dorsal fin and saddle patch, so births, deaths and maternity are recorded rather than estimated across several decades.

How far it can be extended

Life-history parameters differ between ecotypes and populations, and captive figures reflect husbandry rather than the species.

Caveats

  • One population; other ecotypes and regions differ, and some are far less studied.
  • Maximum ages for animals already adult when cataloguing began are estimates.
  • Captive lifespans are a separate question and are not comparable.

Still unanswered

  • How much of the sex difference in longevity is explained by the demands of male size?

Last reviewed 2026-08-11

The evidence (2 studies)

Mating happens between pods rather than within them — a matriline is a group of close relatives, and calves are fathered by males from a different group encountered when pods meet. Males return to their mothers afterwards. There is no pair bond, no territory and nothing that would be recognised as a mating season in most populations, though births cluster loosely.

Gestation runs about seventeen months, among the longest of any mammal, and produces a single calf around 2.5 metres long. The calf nurses on milk delivered under pressure from concealed mammary slits — the mother squirts it into the calf’s mouth, because a calf cannot form a seal underwater — and that milk is roughly 40% fat, closer to butter than to anything a land mammal produces. Nursing continues for a year or two, with solid food introduced well before it ends.

  • First calf at around 14 years, with roughly five years between births.
  • Around half of calves do not survive their first year in the populations where this has been measured.
  • Females stop reproducing in their late thirties or forties and frequently live decades beyond it.
  • Sons stay with their mothers for life. So do daughters. Nothing disperses.
  • Males average around 30 years and reach about 60; females average around 50 and reach into the eighties.

The sons-stay-forever pattern is not sentiment, and the data are unusually direct about it. When a mother dies, her adult son’s risk of dying in the following year rises sharply — by a large factor for a male over thirty — and the effect on daughters is much weaker. Older females lead the group to food in lean years and share what they catch preferentially with their sons. A post-reproductive female is not a spare animal; she is carrying the group’s accumulated knowledge of where the fish are.

What eventually kills an orca is usually invisible. They have no natural predators, and the causes are prey scarcity, disease, injury and accumulated pollutants — PCBs in particular, which concentrate up the food chain into blubber and pass to a first calf in milk. In several populations first calves disproportionately fail, and contaminant load is a leading explanation.

Animal welfare

On captive lifespans

Figures for orcas in marine parks describe husbandry rather than biology and are not comparable with wild demography. NatureHQ does not publish them as facts about the species.

Where this applies: Keeping cetaceans in captivity is restricted or prohibited in a growing number of jurisdictions and permitted in others.

Words used here
Matriline
A female and all her descendants, travelling together permanently. The basic unit of orca society.
Post-reproductive lifespan
Living substantially beyond the last birth. Documented in humans, orcas and a handful of other whales, and almost nowhere else.

An orca is a dolphin, a whale and a mammal all at once, and the three answers only sound contradictory because "whale" is doing two jobs. Biologically, orcas are the largest members of the dolphin family, Delphinidae. Delphinidae sits inside the toothed whales, which sit inside the cetaceans — so every dolphin is a whale in the taxonomic sense, and "whale" in everyday speech simply means a large cetacean that is not called a dolphin. Both usages are defensible; they answer different questions.

What an orca is
QuestionAnswerWhy
Is it a mammal?YesWarm-blooded, air-breathing, gives birth to live young, produces milk
Is it a fish?NoDescended from hoofed land mammals; its closest living relatives are hippos
Is it a dolphin?YesThe largest member of the dolphin family, Delphinidae
Is it a whale?Yes, and also noA toothed whale by descent; not a "whale" in the everyday sense that excludes dolphins
Why "killer whale"?A translation that flippedSpanish whalers called them whale killers, from watching them hunt whales. The words swapped over

Vision is good in air and in water — an orca can watch a person at the side of a pool, and spot a seal on an ice floe from the surface — but it is not the primary sense. Hearing is. Sound travels roughly four and a half times faster in water than in air and carries far further, and an orca receives it through fat-filled channels in the lower jaw that conduct vibration directly to the ear bones. There is no external ear opening worth the name.

Fish-eating populations echolocate constantly, producing clicks in the nasal passages and focusing them with the melon, the fatty organ in the forehead. Mammal-eating populations echolocate much less, and the reason is that seals and porpoises can hear the clicks — the same equipment that finds a herring announces the hunter to a seal. They hunt in near silence and listen instead.

Smell is essentially gone. Toothed whales have lost the olfactory apparatus almost entirely, which makes sense for an animal that opens its nostril at the surface for a fraction of a second at a time.

Words used here
Delphinidae
The oceanic dolphin family. Orcas are its largest members, at up to nine metres.
Melon
The fatty organ in a toothed whale’s forehead that focuses outgoing echolocation clicks.
Echolocation
Building a picture of the surroundings from the echoes of self-produced clicks.

There is no reliable record of a wild orca killing a person. For an eight-metre predator that hunts great white sharks, kills blue whales and encounters small boats routinely, that is a genuinely extraordinary fact, and it is not explained by lack of opportunity. Wild orcas approach vessels, follow swimmers and investigate people in the water, and then do not attack them.

Captivity is a different record entirely. Orcas in marine parks have killed and seriously injured trainers on several occasions. Whatever produces those events — confinement, social disruption, the absence of anything to do — it is a fact about the conditions rather than about the species, and reading it as evidence of what a wild orca will do gets the causation backwards.

Since 2020 a subpopulation off Iberia has been damaging the rudders of sailing yachts, in hundreds of incidents, sinking a small number. Nobody has been hurt. The behaviour spread through the group in a way consistent with social learning, and explanations range from a response to an earlier injury to play — a fad in a population that has invented and dropped fads before, most famously the Pacific Northwest orcas who spent part of 1987 carrying dead salmon on their heads.

Safety

Near orcas

Keep a distance and let the animals set the terms of any encounter. Approaching, feeding or attempting to swim with wild orcas is restricted or illegal in most places they occur, and off Iberia sailors should follow current national guidance for the area rather than any general rule.

Where this applies: Minimum approach distances and permitted activities are set nationally and vary considerably.

When to get help: Consult the relevant maritime authority for current guidance where interactions with vessels are occurring.

  • Are the ecotypes separate species?

    Why it matters: They do not interbreed, differ in diet, society and voice, and a single conservation assessment for all of them is close to meaningless.

    What would settle it: Continued genomic work; several proposals to split them are already under formal consideration.

  • How is ecological knowledge passed from an old female to her group?

    Why it matters: The leadership result shows the knowledge has value. Nobody has observed the transfer.

    What would settle it: Fine-grained movement data linking individual decisions to subsequent group behaviour.

  • Why are orcas interacting with boat rudders off Iberia?

    Why it matters: It is the most-reported orca story of recent years and is almost universally explained by motives nobody has evidence for.

    What would settle it: Longer observation of the identified individuals, and whether the behaviour spreads or fades.

Claims about this, checked

Things people have heard, and what the evidence actually supports.

The research behind this page

11 studies, newest first. Each one has a page explaining what it found and what it could not show.

What this page is still missing

NatureHQ publishes its own gaps. This record is at 82% completeness against what we would call a finished subject.

  • 6 high-priority search intent(s) not yet covered
  • no research from the last few years is attached — check for newer work
  • more experiments could be explained in plain English
  • Transient and offshore ecotypes are described but have no research attached.
  • Captivity has a substantial literature and a serious welfare dimension that this record does not yet cover.

Last reviewed 2026-08-09 · 9 claims · 78 search questions answered on this page