Keystone Species & Trophic Cascade: Otters, Urchins and Kelp

A term coined from a single removal experiment

The phrase keystone species comes from zoologist Robert Paine's 1969 experiments on a rocky Pacific shoreline, where he physically removed the predatory sea star Pisaster ochraceus from test plots and simply watched what happened. Without the sea star preying on them, mussels — normally kept in check — multiplied unchecked and monopolized the rock surface, crowding out over a dozen other species that had coexisted in the predator's presence. Species diversity in the plot collapsed from more than fifteen species to essentially one. The sea star wasn't the most abundant organism in the system by biomass; it was the one whose removal mattered disproportionately, which is the whole point of the term — a keystone, in the architectural sense, is the single central stone that holds an arch together even though it's no bigger than any other stone in it.

Sea otters, urchins and kelp: the canonical modern case

The most cited living example of a trophic cascade is the North Pacific kelp forest, structured around three trophic levels: kelp (the primary producer, a fast-growing brown alga forming underwater forests), sea urchins (herbivorous grazers that eat kelp holdfasts and stipes), and sea otters (the apex predator, whose diet is dominated by urchins). Historical fur-trade hunting nearly eliminated otters across huge stretches of coastline in the eighteenth and nineteenth centuries, and in area after area the pattern repeated: without otter predation, urchin populations exploded, their grazing pressure shifted from patchy and sustainable to dense and relentless, and kelp forests were reduced to urchin barrens — bare rock studded with urchins and almost nothing else standing.

The mechanism: top-down control, not resource limitation

The cascade works because predation is top-down control: the otter doesn't affect kelp directly at all — it never eats kelp — but its effect on urchin abundance and behaviour ripples down a level to shape the primary producer. This is the defining feature that separates a trophic cascade from ordinary resource competition, where species at the same level compete for the same limited food or space. Remove the top predator and the effect isn't confined to its immediate prey; it propagates an extra step to whatever that prey eats, which is why ecologists describe cascades as running two, three, even four levels deep in well-studied systems, with alternating sign at each level — fewer otters means more urchins means less kelp.

// simplified three-level Lotka-Volterra-style cascade
dKelp/dt   =  rK * Kelp * (1 - Kelp/K_carry)  -  g * Urchin * Kelp
dUrchin/dt =  e * g * Urchin * Kelp            -  m * Urchin  -  p * Otter * Urchin
dOtter/dt  =  (birth - deathRate) * Otter       // near-static on the timescale of a cascade

// remove otters → p*Otter*Urchin term vanishes → Urchin grows unchecked
// unchecked Urchin drives g*Urchin*Kelp above kelp's regrowth rate rK
// Kelp → 0 (urchin barren), a stable alternative state, not just a dip

Why impact isn't proportional to biomass

The defining, counterintuitive feature of a keystone species is that its ecological weight has almost nothing to do with how much of the ecosystem's total biomass or energy flow it represents. Sea otters and Paine's sea stars were never the dominant biomass in their ecosystems by a wide margin — kelp and mussels vastly outweighed them. What makes a species a keystone is a structural property of the food web: it sits at a control point where its actions gate the population of something else that in turn controls a large downstream pool of biomass or diversity. Remove a numerically dominant but structurally ordinary species and you usually get a proportional, absorbable gap other species fill; remove a keystone and the whole architecture can flip to a different, often much simpler, stable state — an urchin barren instead of a kelp forest is a genuinely different ecosystem, not just a smaller version of the old one.

Recovery is not automatic

Reintroducing or protecting otter populations does allow a cascade to run in reverse, and it has, in documented cases along the Aleutian and California coasts — but the barren state can be surprisingly stubborn. Dense urchin populations can persist even with predators present, because urchins in a barren often survive on drift algae and detritus rather than living kelp, requiring less food than a barren-clearing regrowth phase would suggest, and because otters preferentially hunt where handling costs are lowest, which doesn't always mean the most degraded patches first. Full kelp-forest recovery after keystone predator return has taken anywhere from a few years to multiple decades depending on the site, underlining that trophic cascades are not simple, instantly reversible dials — the system carries memory of which state it's in.

Frequently asked questions

Try it live

The whole three-level system runs live in Keystone Species & Trophic Cascade. Remove the top predator mid-run and watch the cascade unfold down to bare rock — then bring it back and see how sluggishly the kelp recovers.