Skip to content
NatureHQ

A guided journey · 5 stops

One change, many consequences

Take one species out and see how far the effect travels — from a shore where it is measurable to a national park where it is argued about.

The route

  1. Keystone species
  2. Sea stars
  3. Trophic cascades
  4. Beavers
  5. Wolves

A great deal of what ecology knows with confidence comes from taking something away and watching. The design is unglamorous — a marked plot, a matched control, and years of going back — and it is the difference between a story that fits and a result.

This route follows the removal argument from the place it works best to the place it is claimed hardest. It starts on a rocky shore, where a man with a crowbar produced one of the most influential experiments in the field. It ends with wolves, elk and rivers, where no control exists, six links of inference are told as one sentence, and the most testable link failed its test.

The order matters because the shore teaches you what good evidence for this kind of claim looks like. Everything after it can then be measured against that.

  1. Stop 1 of 5. Ecology

    Start with the word, and the test inside it

    Because the concept has an abundance criterion that almost everybody drops. A keystone species has an effect out of proportion to how much of it there is — so a dominant tree is not one, however important, and a top predator is not one automatically. Getting that straight first stops the rest of the route reading as a general claim that big animals matter.

    Read Keystone speciesNot the biggest stone in the arch. The one whose removal brings it down.

  2. Stop 2 of 5. Marine life

    Then the shore where the idea came from

    This is what the strongest version of the evidence looks like: one plot cleared by hand for years, one untouched beside it, species number halving on the cleared one. It also corrects the mechanism most people assume. The sea star is not controlling numbers — it is eating the best competitor for rock space, so the space stays open. That is a specific enough claim to fail elsewhere, and it does.

    Read Sea starsIt eats a mussel by pushing its own stomach into the gap between the shells.

  3. Stop 3 of 5. Ecology

    Follow the effect one link further

    A cascade is what happens when the effect passes the animal the predator eats and reaches the plants. Read after the shore, the pattern is visible: the evidence is best in ponds, streams and kelp — systems small enough to manipulate — and the claims are loudest about national parks. The otter and urchin case sits between the two, as a natural experiment history ran without asking.

    Read Trophic cascadesThe evidence for cascades is best in ponds and shores. The claims are loudest about national parks.

  4. Stop 4 of 5. Mammals

    Meet the animal the famous story leaves out

    Before the wolves, the beavers. Trapping removed them from most of a continent, and their dams had been holding water tables high across whole catchments. This step is placed before the Yellowstone page deliberately: knowing that the hydrology was already gone changes what the wolf story can plausibly claim, and it is the piece the viral version has no room for.

    Read BeaversThe dam is not the house. It exists to keep the front door under water.

  5. Stop 5 of 5. Mammals

    And end where the evidence gets hardest

    One reintroduction, one park, no control, and a decade in which drought, hunting, bears and a hard winter also acted. Some links are well evidenced. The most testable one — that fear redistributed elk browsing — was tested at landscape scale and not found. And when willow growth was manipulated directly, reduced browsing alone was not enough: the water tables had to come back too. The route ends here because this is where everything the shore taught actually gets used.

    Read WolvesThere is no alpha wolf. The man who put the idea into circulation spent the next thirty years trying to take it out.

Where this leaves you

You end able to ask a specific question of any claim that removing a species changed an ecosystem: what was the comparison? A matched plot with the species still on it? Another island where history left it in place? Or the same place before, with a decade of other changes in between?

The uncomfortable stop is the last. Yellowstone is not a case of anybody lying — elk declined, vegetation recovered in places, beavers increased — and the chain that got compressed into one sentence has links of very different strength. The willow experiment introduces a factor the story has no room for: the beavers had gone a century earlier, and the water tables went with them.

What the route does not settle: how strong cascades are in large terrestrial systems generally, which is one of the field's open questions rather than something this site can resolve.

Last reviewed 2026-09-04. All guided journeys