Species turnover measures how fast the members of an ecological community replace each other — not total biodiversity, not extinction rate, but the rate of substitution. One species leaves; another arrives. The replacement keeps the community dynamic even when the count stays roughly the same. High turnover means the system is responding to change. Low turnover means it isn't — or can't.
Nwankwo and Rossberg (Nature Communications, February 2026) analyzed a century of biodiversity surveys across marine, freshwater, and terrestrial ecosystems and found that turnover rates have declined by roughly one-third since the 1970s, measured over one-to-five-year windows. This is the opposite of what climate models predict. Accelerating climate change should increase ecological pressure, forcing faster species replacement as conditions shift and thermal niches migrate. The turnover rate should be climbing. It's falling.
The mechanism is not that ecosystems have stabilized. It's that they've run out of replacement candidates. Regional species pools — the reservoir of species available to colonize a community when a member leaves — have been depleted by habitat loss, fragmentation, and local extirpation. The engine of turnover is still receiving the same climatic pressure, but the fuel supply has been cut. A community that loses a heat-sensitive species to warming would normally gain a heat-tolerant species from the surrounding region. If that surrounding region no longer contains the heat-tolerant species — because its habitat was converted to agriculture, or because it went locally extinct from a different stressor — the slot stays empty, and turnover drops. The community isn't adapting. It's stalling.
The through-claim is in the ambiguity. A declining turnover rate admits two interpretations, and the data alone can't distinguish them. One: the system is reaching equilibrium, settling into a new stable configuration as communities adjust to changed conditions. Two: the system is losing its capacity to respond, slowing down because the pool of possible responses has been drained. Both produce the same measurement — lower turnover — from opposite causes. Stability and stagnation are indistinguishable from outside the system. You need the second variable. Without the species pool data, a declining turnover rate looks like a system finding its footing. With the species pool data, it looks like a system losing its legs.
This is not unique to ecology. Any system that responds to perturbation by drawing from a finite reservoir will show declining response rates as the reservoir depletes — and the declining rate will look like adaptation until you measure the reservoir directly. The diagnostic is never the response rate alone. It's the response rate divided by what's available to respond with.