Skip to content
NatureHQ

ability

Animal memory

Memory is several separate systems, not one capacity. Some animals remember what happened, where and how long ago; some grow the brain region for spatial memory when they need it; and none has been shown to relive a moment.

Ask which animal has the best memory and the honest answer is that the question does not have one, because memory is not a single thing. A honey bee runs at least four mechanically distinct memory phases in a brain of about a million neurons. A scrub jay recovering its own caches combines what it buried, where, and how long ago — and will abandon its favourite food if enough time has passed for it to have spoiled. A chickadee from a harsher climate grows a larger hippocampus than one from a milder place, within its own lifetime. Those are three different capacities, and an animal can be exceptional at one while unremarkable at another.

Developed record · 70% complete · reviewed 2026-08-09

What this page covers

A capability rather than an organism. Covers memory wherever it has been experimentally dissociated — insects, corvids, elephants and cetaceans.

Quick facts

Kind
A capability page, assembled from evidence across species
Systems
Short-, mid- and long-term phases; spatial, associative, social and episodic-like
Episodic-like
Demonstrated in scrub jays — what, where and when, not recollection
Not demonstrated
Human-style autobiographical recall, in any species

Where this appears

Assembled from the knowledge graph. Each entry carries its own evidence and its own limits.

Memory is not one thing

Memory is not one thing animals have more or less of — it is several separate systems

Established

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

Distinguishable memory systems with different durations, capacities and mechanisms have been demonstrated within single species: honey bees show short-, mid- and long-term phases separable by protein-synthesis dependence, while spatial, associative and social recognition memory dissociate across tasks in birds and mammals.

Who this applies to
the species in which memory systems have been experimentally dissociated
Studied in
Apis mellifera, Aphelocoma californica, Poecile atricapillus
Why we rate it this way, and what the caveats are
EstablishedHigh confidence

Memory phases in bees are separable pharmacologically, and spatial memory dissociates from other kinds by lesion and by task across birds and mammals. The finding is old, replicated and mechanistically grounded.

How far it can be extended

Assembled from primary work in each species. The claim is about the architecture of memory rather than about any one animal, so its strength comes from the same organisation appearing in brains of very different sizes.

Caveats

  • Ranking species by "memory" is close to meaningless — the question is always which memory, for what.
  • Phase boundaries are defined by the interventions that disrupt them and are not sharp.
  • Most of this work uses narrow laboratory tasks.

Still unanswered

  • How far do the memory phases described in insects map onto vertebrate systems?

Last reviewed 2026-08-09

The evidence (2 studies)

The systems worth keeping apart are these. **Working memory** holds something for seconds while you use it. **Associative memory** links a cue to an outcome — the foundation of nearly all animal training. **Spatial memory** stores places, and is the one caching animals develop to extraordinary degrees. **Social recognition** stores individuals and, in some species, the relationship attached to them. **Episodic-like memory** binds content, place and time into one record.

These dissociate: an animal can be outstanding at one and ordinary at another, and damage or interference can remove one while leaving the rest. That is why "which animal has the best memory" cannot be answered as asked, and why any list ranking species by memory is measuring nothing in particular.

Words used here
Episodic-like memory
Remembering what happened, where, and how long ago, as one integrated record. The hyphenated "-like" is doing real work: it marks that nobody has shown the animal re-experiences the event.
Consolidation
The process by which a fresh, fragile memory becomes a durable one — in bees, marked by a switch to requiring new protein synthesis.

Remembering what, where and when

Some animals remember what they did, where, and how long ago — but that is not the same as remembering the moment

Well supported

Good evidence backs this, though some details remain open.

Scrub jays recovering their own caches integrate the identity of the food, its location and elapsed time, switching from preferred perishable prey to durable prey after longer delays. This satisfies the behavioural criteria for episodic-like memory without demonstrating subjective recollection.

Who this applies to
western scrub jays in cache-recovery experiments
Studied in
Aphelocoma californica
Why we rate it this way, and what the caveats are
Well supportedHigh confidence

A clean design in which the birds cached the food themselves and the only way to behave correctly was to combine all three elements. Replicated and extended over two decades.

Caveats

  • The authors said "episodic-like" deliberately. Coverage drops the qualifier, and dropping it changes the claim.
  • No behavioural design can show whether an animal re-experiences an event, which is what episodic memory means in humans.
  • Whether jays track elapsed time or read a decaying memory trace is not fully separated.

Still unanswered

  • Is there any way to test for subjective recollection in a non-verbal animal?
  • How widespread is what-where-when memory outside caching species?

Last reviewed 2026-08-09

The evidence (1 study)

The design is what makes this convincing. The jays cached the food themselves, so nobody could have cued them; the two foods differed in how fast they spoil; and the only way to behave correctly after a long delay was to combine all three elements — which food, in which place, how long ago. A bird tracking only location, or only preference, fails.

What it does not show is the part everyone wants it to. Episodic memory in humans involves re-experiencing an event, and no behavioural experiment can reach that in an animal that cannot be asked. Hence "episodic-like", a qualifier the original authors chose carefully and popular coverage almost always drops.

Where these findings live

  • Common raven

    Remembering individuals, and the relationship, for years

  • Elephants

    Matriarchs as repositories of knowledge the group depends on

  • Honey bee

    Several memory phases in a very small brain

The same test, in a bird and in an ape

Apes remember what was hidden, where, and how long ago — and act on all three

Well supported

Good evidence backs this, though some details remain open.

Great apes shown food cached in two locations — one preferred but perishable, one durable — choose the perishable option after short delays and the durable option after long ones, integrating content, location and elapsed time in a single retrieval decision.

Who this applies to
captive great apes across three genera
Studied in
Pan troglodytes, Pongo pygmaeus, Gorilla gorilla
Why we rate it this way, and what the caveats are
Well supportedModerate confidence

The design is the established behavioural criterion and the result replicates across genera. Confidence is held at moderate because the criterion cannot distinguish remembering an event from holding a structured record of one.

How far it can be extended

The what-where-when criterion has been met independently in chimpanzees, orangutans and gorillas, and separately in scrub jays.

Caveats

  • Small numbers of captive, heavily tested individuals.
  • The "-like" in episodic-like is doing real work: subjective recollection is not testable behaviourally.
  • Elapsed-time tracking may use an interval timer rather than a memory of when.

Still unanswered

  • Is there any behavioural test that could distinguish recollection from a structured record?
  • How widely does what-where-when memory occur outside apes and corvids?

Last reviewed 2026-08-09

The evidence (2 studies)

How we know

The jay that remembers which cache has gone off

Can a bird remember not just what it hid and where, but how long ago — and act on all three at once?

Scrub jays were allowed to cache two foods: worms, which they much prefer but which decay, and peanuts, which keep. They were then allowed to recover the caches after either a short delay or a long one. An animal remembering only what and where should always go for the preferred worms. An animal that also tracks elapsed time should switch to peanuts once the worms have had time to spoil.

What happened

After short delays the jays went to the worms. After long ones they switched to the peanuts — integrating what was cached, where, and how long ago into a single retrieval decision.

What it shows

The what-where-when criterion is met in a bird, a decade before it was demonstrated in great apes. Whatever this capacity is, it is not a primate inheritance that birds happen to share; it has appeared at least twice.

What it does not show

This is why the literature says "episodic-like". A behavioural test cannot distinguish remembering an event from holding a structured record of one, and no experiment of this kind can establish that an animal re-experiences its past. Elapsed time might also be tracked by an interval timer rather than a memory of when.

The controls — what makes this evidence rather than a story
  • Both foods were cached by the same bird in the same session, so preference and location are held constant and only elapsed time varies.
  • Birds with no experience of degraded worms provided the comparison, testing whether the switch is learned from the food going off.
  • Cache sites were distinct and trays were replaced, preventing recovery by smell or residual cues.

From Episodic-like memory during cache recovery by scrub jays

Chimpanzees, orangutans and gorillas watch food being hidden in two places — one preferred but perishable, one durable — and are let back in after a short or a long delay. They go for the good perishable food when little time has passed and switch to the durable option when a lot has. Content, place and elapsed time, integrated in a single choice.

The scrub jays got there first, by a decade, which is the point worth holding onto: the capacity is not a primate inheritance that birds happen to share. It has appeared at least twice.

The hedge in "episodic-like" is doing real work and NatureHQ keeps it. A behavioural test cannot separate remembering an event from holding a structured record of one, in an ape or in a jay — and asserting that an animal re-experiences its past is not something any experiment of this kind can establish.

Memory grows where it is needed

Animals that need better memory grow the brain region for it — within a single lifetime

Well supported

Good evidence backs this, though some details remain open.

Food-caching chickadees from harsher environments, and birds held on restricted food, show larger hippocampi with more neurons and better spatial memory performance than conspecifics facing lower demand.

Who this applies to
black-capped chickadees across populations and food regimes
Studied in
Poecile atricapillus
Why we rate it this way, and what the caveats are
Well supportedModerate confidence

The population comparison is correlational, but the food-restriction manipulation supplies an experimental component. Confidence is held at moderate because hippocampal volume is a coarse proxy for memory capacity.

How far it can be extended

Hippocampal responsiveness to spatial demand has been reported across several food-caching bird families; it was measured here in one species.

Caveats

  • Volume is a blunt measure; more neurons is not automatically more memory.
  • Population differences confound genetics with experience, which the food manipulation only partly separates.
  • Caching strategies differ between species, so this does not transfer uniformly.

Still unanswered

  • Does the same plasticity operate in non-caching animals facing spatial demand?

Last reviewed 2026-08-09

The evidence (1 study)

This is the finding that most changes how the subject should be read. Memory capacity is not a fixed species property sitting somewhere on a ladder. A chickadee facing a hard winter, which must relocate thousands of caches to survive it, develops more of the brain region that does the work than a chickadee that does not — and the difference is partly produced by the conditions rather than inherited.

It also explains the familiar claim that caching birds have astonishing spatial memory. They do. The reason is that they would die without it, and the capacity tracks the need closely enough to vary between populations of the same species.

Words used here
Hippocampus
A brain region central to spatial memory in birds and mammals. In caching birds its size varies with how much the animal needs it.

What we do not know

  • Can subjective recollection be tested in a non-verbal animal at all?

    Why it matters: It is the difference between episodic-like and episodic memory, and the whole popular version of this subject rests on the distinction being ignored.

    What would settle it: A behavioural signature that recollection produces and reconstruction cannot. Several have been proposed; none is accepted.

  • How long do animal memories actually last in the wild?

    Why it matters: Almost every duration figure comes from captivity, where nothing competes for attention and the animal is tested repeatedly.

    What would settle it: Long-term individual tracking with natural intervals, as the raven playback work began to do.

  • Do the memory phases in insects map onto vertebrate systems?

    Why it matters: If a bee and a mammal organise memory the same way, that organisation is either very old or a very strong attractor.

    What would settle it: Comparative molecular work on consolidation across distant lineages.

The research behind this page

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

You may also find this interesting

What this page is still missing

NatureHQ publishes its own gaps. This record is at 70% completeness against what we would call a finished subject.

  • 1 high-priority search intent(s) not yet covered
  • no research from the last few years is attached — check for newer work
  • no popular claim about this subject has been checked yet
  • Octopus and cephalopod memory is absent, despite being one of the most-asked areas.
  • Working memory has no primary research attached.
  • Human comparison is deliberately avoided but never explained; the page would be stronger for saying why.

Last reviewed 2026-08-09 · 6 claims · 6 search questions answered on this page