Not every species matters equally to the ecosystem it lives in. Some species could vanish and leave the surrounding web of relationships largely intact; others, once removed, can trigger a cascade of secondary extinctions and structural collapse far larger than their own numbers would predict. Ecologists call this second kind a keystone species — named, deliberately, after the wedge-shaped stone at the top of an arch that holds the entire structure up despite bearing no more individual weight than any other stone in it.
Why abundance isn't what makes a species a keystone
A keystone species is defined by its disproportionate effect on ecosystem structure relative to its actual biomass or population size, not by how numerous or dominant it is. The classic demonstrated example is the sea otter in North Pacific kelp forest ecosystems: otters prey on sea urchins, and where otter populations were reduced by hunting, urchin populations exploded and grazed kelp forests down to bare rock, a state ecologists call an urchin barren, eliminating the habitat structure that supported a much wider community of fish and invertebrates. Reintroducing otters to areas where they'd been eliminated has been observed to allow kelp forests to recover, with urchin populations brought back under control by predation — a single mid-level predator, never especially numerous itself, turning out to control the entire ecosystem's basic structure through its effect on one prey species further down the food web.
Removing one thread from a web whose structure you can't fully see in advance
Identifying which species in a given ecosystem functions as a keystone generally isn't obvious in advance from population counts or biomass alone; it typically becomes clear only through careful, often long-running ecological study, or in the worst case, only after that species has already been removed and the resulting cascade of changes has played out. This is part of why conservation biologists treat keystone species identification as a priority wherever it's feasible: knowing in advance which relatively small number of species are disproportionately load-bearing for a given ecosystem lets conservation effort be targeted at protecting those species specifically, rather than spreading limited resources evenly across every species present, on the reasoning that losing a keystone species risks a structural collapse that losing a comparably rare non-keystone species would not.
What we're still unsure about
The keystone species concept itself, and well-studied cases like sea otters and Pacific kelp forests, are well established in ecology through decades of field research. What's harder to pin down with confidence is identifying keystone species in ecosystems that haven't been as intensively studied, since the concept depends on understanding a whole web of indirect interactions that isn't always visible from short-term observation, and ecologists continue to debate how universally the concept applies versus how much any given ecosystem's structural dependencies are genuinely unique to it and resistant to generalisation from other, better-studied cases.
This sits inside Biodiversity & Conservation Biology, one of seven topics in Ecology, one of six domains in Biology, one of seventeen subjects the app can quiz you on.