In 1963, the marine ecologist Robert Paine picked up every starfish — Pisaster ochraceus — from a 25-meter section of intertidal rocks at Mukkaw Bay, Washington, and threw them back into the sea. He repeated the removal each month, year after year. The control plot beside his ran undisturbed. After a few seasons the difference was unmistakable: in the control, fifteen species of mussels, barnacles, chitons, limpets, and algae coexisted at modest density. In the Pisaster-removal plot, mussels — Pisaster's preferred prey — outcompeted everything else and the community collapsed to a near-monoculture. The starfish, present at only a few individuals per square meter, was somehow holding the entire community open. Paine called this kind of organism a keystone species — the stone in the arch whose removal brings the rest down — and the field of experimental community ecology dates from this paper.
A trophic cascade is what happens when the keystone is at the top of the food web. Remove the apex predator and herbivore populations expand; herbivore populations expand and primary producers collapse; primary producers collapse and the system reorganizes in ways the original predator was indirectly preventing. The most-cited example is the gray wolf reintroduction to Yellowstone in 1995. Wolves changed elk grazing behavior — pushing elk away from open valleys onto thicker forest edges — which let willows and aspens recover along stream banks; recovering willows let beavers re-establish; beaver dams restored wetlands. The popular version of the story has been challenged as overstated (the magnitude of the effect varies across the park, and the willow recovery has confounders), but the form of the argument has been observed in many systems. Sea otters reduce sea-urchin grazing in California kelp forests; reef sharks on Caribbean reefs reduce mid-level predator pressure on herbivorous fish, which keeps coral surfaces clear of algae.
The mirror phenomenon is mesopredator release. When top predators disappear, the middle predators they previously suppressed — coyotes, raccoons, feral cats, smaller hawks — expand sharply, and the prey species those predators target are disproportionately affected. The Pleistocene megafauna extinctions — mammoths, mastodons, giant sloths, sabertooth cats, dire wolves, all gone within the last twelve thousand years — are read in this framework as the largest natural experiment in apex-predator and apex-herbivore loss the planet has run, and the resulting reorganization of plant communities, fire regimes, and nutrient distribution is still being mapped. Apex predator restoration is now a deliberate conservation tool, with reintroductions of wolves, lynx, condors, and beavers underway in dozens of programs globally — on the assumption that the cascade runs the other way too, and that adding the keystone back can sometimes restore the structure.
The cascade logic lands hardest in fisheries. Large predatory fish — cod, tuna, sharks — have been reduced to roughly 10% of their pre-industrial biomass across the world's oceans. The downstream consequences are being measured in real time: overpopulation of mid-level fish and squid, expansion of jellyfish blooms, shifts in plankton communities. The cod fishery of the Grand Banks, off Newfoundland, collapsed in 1992 after centuries of intensive fishing and has not recovered in the three decades since — the canonical example of a cascade that does not reverse on the timescale humans planned for. The keystone species concept, born in a tide pool in the 1960s, now reads as a description of the structural state of the global ocean.