Skip to content
NatureHQ

Ecologyphenomenon

Keystone species

Not the biggest stone in the arch. The one whose removal brings it down.

A species whose effect on its community is out of proportion to how much of it there is. The disproportion is the whole point: a species that matters exactly as much as its abundance suggests is dominant, not keystone.

The word came from a short note in 1969, written after several years of prising sea stars off a stretch of rocky shore. The analogy is the stone at the top of an arch: not the largest, and the one whose removal brings the structure down. What the note proposed was specific and testable — that some species influence a community out of proportion to their numbers — and what happened to the word afterwards is a case study in how a technical term escapes. Within two decades keystone was being applied to whales, to elephants, to dominant trees, to any species someone considered important, including the most abundant ones in their systems. That is nearly the reverse of the original meaning. By 1996 a group of ecologists including the field’s leading figures published a paper effectively asking for the word back, and the test they proposed is the one this site uses: impact measured relative to abundance, against a stated baseline. If a species is as influential as its abundance would predict, it is not a keystone. It is common and important, which is a different claim and a much easier one to make. Two further distinctions are worth holding. A keystone is not the same as an apex predator: a top predator may be a keystone, and may equally be one of many pathways in a food web whose removal changes little. And a keystone is not an ecosystem engineer — a beaver changes its surroundings by building, which is a different mechanism from eating a competitive dominant, and the two words are frequently used as if interchangeable. Finally, and most importantly, keystone status is a property of a situation rather than of a species. The same sea star is a keystone on a shore where mussels would otherwise monopolise the rock, and considerably less consequential on a shore where they would not.

Early coverage · 66% complete · reviewed 2026-09-04

What this page covers

Keystone status belongs to a species in a particular community, not to the species everywhere. Demonstrated cases are far fewer than claimed ones, because demonstrating it requires a removal experiment.

Often confused with: An important species, which is a much weaker and much commoner claim; An apex predator, which may or may not be a keystone; An ecosystem engineer, which acts by changing the physical environment instead; An abundant or dominant species, which is close to the opposite of the definition

Quick facts

The test
Large effect relative to abundance — not just a large effect
Not apex predator
A top predator may be a keystone, or may be one pathway among many
Not ecosystem engineer
Engineers act by changing physical structure, which is a different mechanism
It belongs to the situation
The same species can be a keystone on one shore and unremarkable on another

A word that escaped its evidence

Coined carefully in 1969; applied to almost anything by 1990.

A keystone species has an effect out of proportion to how much of it there is. That "out of proportion" is the whole content — an abundant, obviously important species is not a keystone, just dominant.

Established

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

Keystone status denotes community impact large relative to the species’ abundance or biomass. Establishing it requires measuring impact per unit abundance against a stated baseline, rather than demonstrating that removal produces change.

Who this applies to
A definition applied across communities; demonstrated designations are far fewer than claimed ones.Do not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Animalia, Plantae
Why we rate it this way, and what the caveats are
EstablishedHigh confidence

The definition and its abundance criterion are settled in the literature; what is contested is which species meet it.

How far it can be extended

Keystone status is a property of a species in a particular community, not of the species everywhere.

Caveats

  • Most species called keystones have never had their impact measured per unit abundance.
  • A species can be a keystone in one community and unremarkable in another; the label belongs to the situation.

Still unanswered

  • How many published keystone designations would survive a proper quantitative test — largely unexamined.

Last reviewed 2026-09-04

The evidence (2 studies)
Four things "keystone" gets used for, and what each actually means
Called a keystoneWhat it usually isIs it a keystone?
A sea star at low density holding back musselsLarge effect, low abundanceYes — this is the original case
A dominant forest treeLarge effect, very high abundanceNo — a foundation species instead
An apex predatorSometimes disproportionate, often one route among manyOnly if measured and shown
A beaverChanges the physical environment by buildingAn ecosystem engineer; may also be a keystone

Diagram

Effect against abundance

The distance above the line is the claim. Height alone is not.

Effect against abundance — which is the whole testEffectAbundanceEffect as abundance predictsKeystoneRare, huge effectDominantAbundant, big effect —exactly as expectedDistance above the dashed line is the claim. Height alone is not.
The same explanation in words

A plot with abundance along the horizontal axis and effect on the vertical. A dashed diagonal marks the effect a species would have if it mattered exactly as much as its abundance predicts. A keystone sits high above that line at low abundance — rare, with a large effect. A dominant species sits high on the plot but on the line: abundant, with a big effect, exactly as expected. A note records that distance above the dashed line is the claim, and height alone is not.

The second row is the one that does most damage, because it inverts the definition entirely. A species that dominates a community by being overwhelmingly abundant is important, and its removal would change everything — but its effect is exactly what its abundance predicts. The word was invented for the opposite situation.

Where the idea came from

A crowbar, a stretch of rock, and several years of going back.

Take one sea star off a patch of rocky shore and keep it off, and mussels take over the rock. Species number fell from about fifteen to about eight — while the untouched plot beside it did not change.

Established

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

Experimental exclusion of Pisaster ochraceus from an intertidal plot over several years resulted in competitive monopolisation of primary space by Mytilus californianus and a decline in species richness from roughly fifteen to eight, relative to an adjacent unmanipulated control.

Who this applies to
One rocky shore system on the Pacific coast of North America.Do not extend this beyond the taxa listed — the popular version over-reaches.
Studied in
Pisaster ochraceus, Mytilus californianus
Why we rate it this way, and what the caveats are
EstablishedHigh confidence

A sustained manipulation with a matched control and a clear, large effect, repeatedly reproduced in related systems since.

How far it can be extended

A single removal plot and control on one shore, where the limiting resource is rock space — an unusually simple case.

Caveats

  • One shore, one plot, one control — the result is strong and its geographic scope is narrow.
  • Rock space as the limiting resource makes competitive monopolisation unusually direct; systems with more complex limitation may behave differently.

Still unanswered

  • How the same community responds to sea star loss from disease rather than from experimental removal, which affects a far larger area at once.

Last reviewed 2026-09-04

The evidence (1 study)

How we know

Taking the sea stars off the rock, and keeping them off

What does a predator do to the community it hunts in — beyond reducing the numbers of what it eats?

A section of rocky shore was cleared of its predatory sea stars by hand, and they were removed again whenever they returned, for several years. An adjacent comparable stretch of shore was left untouched. On both, the species present and the area each occupied were recorded through time.

What happened

On the cleared plot, mussels spread and progressively took over the rock, and species number fell from around fifteen to about eight. The untouched plot kept its composition.

What it shows

That a predator can maintain the diversity of a community rather than diminish it. Mussels are the best competitor for rock space and will take all of it; the sea star eats them faster than they can monopolise, and everything else lives in the space that leaves. The predator is not controlling numbers so much as preventing a takeover — which is a different mechanism from the one most people picture.

What it does not show

One shore, one plot, one control. It does not show that predators generally increase diversity: where the prey is not a competitive dominant, removing its predator does nothing of the kind, and in other systems predation reduces diversity. Rock space is also an unusually simple limiting resource, and the clarity of the result partly reflects that.

The controls — what makes this evidence rather than a story
  • The adjacent untouched stretch, which is what turns years of change into a comparison rather than a story.
  • Sustained removal rather than a single clearance, so the effect is of the predator’s absence and not of the disturbance.
  • Area occupied recorded as well as presence, so competitive displacement is visible rather than inferred.

From Food web complexity and species diversity

Diagram

One plot cleared, one left alone

The control is what turns years of change into a result.

One plot cleared, one left alone, several yearsControl plotSea stars left in place~15 speciesComposition unchangedRemoval plotSea stars taken off, kept off~8 speciesMussels take the rockThe predator was not controlling numbers. It was preventing a takeover.Mussels win the competition for space; eating them keeps the space open.
The same explanation in words

Two plots side by side. The control plot, with sea stars left in place, holds around fifteen species and its composition is unchanged. The removal plot, marked out as the manipulated one, has sea stars taken off and kept off; it falls to about eight species as mussels take the rock. Beneath, a summary states that the predator was not controlling numbers but preventing a takeover: mussels win the competition for space, and eating them keeps the space open.

The mechanism is worth stating precisely, because the popular version — predators keep prey numbers down, which is good for everyone — is not what happened. Mussels are the best competitor for rock space and would take all of it. The sea star eats mussels faster than they can monopolise, so the space stays open and everything else lives in it. The predator is preventing a takeover rather than controlling a population, and that is a specific enough claim to fail: where the prey is not a competitive dominant, removing its predator does nothing of the kind.

Where this leads

  • Sea stars

    The animal, and what happened when disease did the removal

  • Trophic cascades

    Effects travelling further down the chain

  • Ecology

    Why the removal design carries so much weight

How you would actually establish it

And why so few designations have.

Establishing keystone status requires two things: removing the species and measuring what changes, and comparing that change against how abundant the species was. The second is the step almost always skipped. Without it, "removing this species changed the community" is compatible with the species simply being a large part of the community, which is not a keystone effect at all.

That requirement explains why demonstrated keystones cluster in intertidal, stream and small-lake systems. You cannot remove an elephant population for five years and keep a control, and the species for which the claim is made most often are exactly the ones where the experiment is impossible. This is not a reason to assume such species are not keystones. It is a reason to notice that the claim is usually an inference.

  • How many published keystone designations would survive a quantitative test?

    Why it matters: The concept is used in conservation prioritisation, where designating a species keystone can direct resources. If most designations rest on importance rather than on disproportion, those decisions are resting on a weaker claim than they appear to.

    What would settle it: Systematic re-examination of published designations against impact-per-unit-abundance, which nobody has undertaken at scale.

Claims about this, checked

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

The research behind this page

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

2010Ecology

Are wolves saving Yellowstone’s aspen? A landscape-level test of a behaviorally mediated trophic cascade

Aspen were not recovering preferentially in high-risk areas.

2005Frontiers in Ecology and the Environment

Loss of foundation species: consequences for the structure and dynamics of forested ecosystems

Loss of a structurally dominant species changes light, temperature, soil moisture, streamflow and nutrient dynamics, and the associated communities do not simply transfer to the replacing species.

1996BioScience

Challenges in the quest for keystones

The term had been applied to species of every abundance and to effects of every size, including cases where the species was simply dominant.

1994Oikos

Organisms as ecosystem engineers

A large class of ecological effects operates through physical modification of the environment rather than through consumption, and these effects are distinct in mechanism from trophic ones.

1974Science

Sea otters: their role in structuring nearshore communities

Islands with otters had few and small urchins and extensive kelp.

1969The American Naturalist

A note on trophic complexity and community stability

Some species have effects on community structure disproportionate to their numbers or biomass, so that their removal changes the community more than their abundance would suggest.

1966The American Naturalist

Food web complexity and species diversity

On the removal plot, mussels expanded and progressively monopolised the rock.

This page is a stop on a longer route

A guided journey reads several subjects in a deliberate order, with an argument for why one follows another. You can join in the middle.

Where to go from here

Each of these follows from something on this page — a relationship in the evidence, a claim people ask about, or the next mechanism along.

How complete this page is, and what it is still missing

NatureHQ publishes its own gaps. This page is at 66% completeness against what we would call a finished subject, and was last reviewed on 2026-09-04. It carries 6 claims and answers 12 mapped search questions.

  • no research from the last few years is attached — check for newer work
  • more experiments could be explained in plain English
  • Foundation species — abundant habitat-formers — are distinguished here but covered on their own page.
  • Keystone mutualists and keystone prey, both proposed variants, are not covered.
  • Conservation use of the concept is noted but not examined.