The phrase "keystone species" gets used a lot in nature writing, and it gets used loosely. People apply it to charismatic animals (wolves, otters), to plants they personally like (oaks, milkweed), to insects they want to defend (honeybees, monarchs). Sometimes the usage is right. Often it isn't. Here's what the term actually means, where it came from, and why oak qualifies but honeybees don't.
The phrase was coined by an American ecologist named Robert Paine in 1969. Paine was working in tide pools on the Pacific coast, and he ran an experiment that changed how ecology is taught. He removed all the starfish (Pisaster ochraceus) from one section of the intertidal zone, and left them in another. Then he watched.
In the section without starfish, the mussels — which the starfish had been eating — spread quickly, covered the rocks, and crowded out everything else. Within a year, the diverse tide-pool community of fifteen or so species had collapsed to a near-monoculture of mussels. In the section with starfish, the community remained diverse, with all fifteen species present.
Paine called the starfish a keystone species, by analogy to a stone arch. Remove the keystone — the wedge-shaped stone at the top — and the arch falls. Remove the starfish, and the entire community collapses. The starfish was disproportionately important. It was holding the structure up.
The technical definition.
A keystone species, in the way Paine and modern ecologists use it, is a species that has a disproportionately large effect on its ecosystem relative to its abundance. It's not the most numerous, the biggest, or the showiest. It's the one whose removal causes the system to fail.
This is a higher bar than people often realize. Most species, even important ones, are not keystone species. If you removed all the goldfinches from a yard, you'd lose goldfinches, but the rest of the food web would mostly continue. Goldfinches are part of the system; they aren't holding it up.
Whereas if you removed all the oaks from an eastern North American forest, you wouldn't just lose oaks — you'd lose the four hundred-plus caterpillar species that depend on them, the migratory birds that depend on those caterpillars to feed their young, the predators that depend on those birds, and so on, all the way up. The system would not just lose a species. It would fail.
What qualifies, what doesn't.
Some genuine keystone species in eastern North American ecosystems:
- Oak (Quercus). Hosts 477+ caterpillar species; supports the entire bird food web in deciduous forests.
- Beavers. One beaver dam creates wetland habitat for hundreds of other species. Remove the beaver, the wetland drains, the community collapses.
- Native milkweed (Asclepias). The only host plant for monarch butterflies. No milkweed, no monarchs.
- Wolves (in their historical range). Suppress deer populations enough to allow tree regeneration; their absence reshapes forests.
Some things that are commonly called keystone species but technically aren't:
- Honeybees. Honeybees are not native to North America. They are managed agricultural pollinators. Removing them would cost commercial agriculture but would not collapse native ecosystems — native bees would (slowly) reclaim their pollination role.
- Monarchs. Beautiful and ecologically meaningful, but not technically keystone — their disappearance would not cause cascading collapse of other species. Their absence would be a tragedy, not a structural failure.
- Most charismatic megafauna. Black bears, deer, hawks — ecologically important, but not keystone in the Paine sense.
Why this matters for your patch.
The reason to care about precision here is that keystone species deserve different treatment than other species. If you have limited space, limited budget, or limited time, the keystone species are the ones to prioritize. Plant the oak before you plant the dogwood. Plant the milkweed before you plant the petunia. The math says these particular plants do disproportionate work.
Robert Paine died in 2016. The starfish on the Pacific coast are currently in serious trouble — a wasting disease has decimated Pisaster populations in the last decade, and the tide-pool communities he studied are visibly collapsing. The keystone metaphor has turned out to be even more accurate than he originally proposed. The arch really does fall.
— The editors