The Switch
Across 360 experimental plots stretching from the snowy Alps to the Negev Desert, evolution worked exactly as predicted. Plants adapted, beneficial genes spread through populations, and survival followed the textbook rules of natural selection. Except in the places that mattered most. In the warmest locations, some populations evolved predictably and survived while others showed seemingly random genetic trajectories that preceded extinction, according to research by the Genomics of rapid Evolution to Novel Environment (GrENE) network consortium. The difference wasn't gradual. It was binary, like a switch flipping off.
Moisés Expósito-Alonso and colleagues had designed the experiment to answer a straightforward question: how fast would plants evolve under different climate stresses? What they discovered instead was that evolution has a breaking point. Beyond a certain threshold of environmental extremity, the engine of adaptation doesn't just slow down. It transforms into something else entirely, a roulette wheel where genetic changes become random noise rather than survival signals.
The Architecture of Observation
The scale of the experiment made the invisible visible. The GrENE network consortium created simultaneous experiments across 30 different climate zones in Western Europe, the Mediterranean, the Middle East, and North America, according to the research. Each location contained 12 separate plots, totaling 360 distinct experiments across three generations of plants. The subject was Arabidopsis thaliana, a genetically diverse annual plant in the mustard family that serves as a standard model organism precisely because its genetic machinery mirrors the fundamental processes shared across plant life.
Genetic diversity matters for evolution because natural selection requires variation to work with. When a population faces new environmental pressures, individuals carrying beneficial genetic variants survive and reproduce more successfully. Those adaptive variants increase in frequency across generations. The population evolves. This is the mechanism that has shaped life on Earth for billions of years, operating on timescales from decades to millennia depending on generation time and selection pressure.
The experiments revealed that evolution can occur at a tempo of three to five years when given sufficient genetic diversity, per the analysis. Researchers directly observed how certain DNA variants, adaptive variants, take over in populations during evolution. In most cases, plants evolved genetically to adapt to their new environments across the first three years of genomic data. The speed was remarkable, but the mechanism was familiar.
The Divergence
Then came the anomaly. Some experimental populations, especially those in the most extreme warm climates, showed no signs of early evolution, according to the findings. But the absence of adaptation wasn't the most striking pattern. In the warmest environments, the data split into two distinct trajectories. Populations with predictable evolutionary changes survived, the research found. Populations with chaotic genetic changes went extinct.
The chaos wasn't metaphorical. Where natural selection should have been filtering genetic variants based on their contribution to survival, the genetic changes appeared random, disconnected from fitness. The populations weren't failing to evolve. They were experiencing something different: genetic drift, the random fluctuation of gene frequencies that dominates when populations become too small for natural selection to maintain its organizing force.
Extreme heat limits populations to small sizes, which can push populations toward an evolutionary breaking point, the analysis revealed. This is the mechanism that transforms evolution from a directional process into randomness. When environmental stress kills enough individuals, the surviving population shrinks below a critical threshold. At that point, which genetic variants persist and which disappear becomes a matter of chance rather than advantage. A beneficial mutation can vanish simply because the individual carrying it dies in a random weather event. A neutral or even slightly harmful variant can spread because it happens to be present in the few survivors.
The Threshold
What the experiment exposed was not that extreme heat kills plants. That's obvious. The revelation was the existence of a specific breaking point where the fundamental nature of evolutionary change transforms. Above the threshold, natural selection functions. Populations adapt. Beneficial genes spread. Survival correlates with fitness. Below the threshold, genetic drift dominates. Changes become random. Adaptation becomes impossible even when genetic diversity exists. Extinction follows.
This rewrites the relationship between climate change and evolution. The standard model assumes that species with sufficient genetic diversity can adapt to new conditions given enough time. The experiment reveals that assumption is dangerously incomplete. Genetic diversity is necessary but not sufficient. Populations must also maintain enough individuals for natural selection to overpower random genetic drift. When extreme conditions shrink populations past the critical size, all the genetic diversity in the world cannot save them because the mechanism of adaptation has stopped functioning.
The goal was to understand how fast plants would evolve under different climate stresses, according to the GrENE consortium. The answer turned out to be more complex than a simple rate. Evolution happens quickly, in three to five years, when populations stay above the threshold. When they fall below it, evolution doesn't happen slowly. It doesn't happen at all. Chaos precedes extinction with equal speed.
The Conservation Paradox
Protected species in natural parks will still suffer from changing local climates, the research indicates. This finding exposes a fundamental flaw in current conservation models. Protecting habitat preserves space but not evolutionary capacity. A nature reserve can maintain pristine conditions while the species within it crosses the evolutionary breaking point, their populations shrinking under climate stress until genetic drift overwhelms selection and extinction becomes inevitable.
The implications cascade through conservation planning. Information about evolution speed and genetic shifts is key to creating models identifying species at risk from environmental change, per the findings. But speed is only half the equation. The new variable is the threshold itself: the population size below which evolution stops working. Species can't be assessed solely on their heat tolerance or genetic diversity. The critical question becomes whether they can maintain populations above the breaking point as climates shift.
The System Redefined
The experiment provides the first direct observation of evolution's off-switch in action. Previous research had documented population declines and extinctions. Other studies had measured adaptation rates. But the simultaneous experiments across 30 climate zones, with 12 plots each, created the resolution needed to see both outcomes, survival and extinction, under comparable conditions and identify the mechanism distinguishing them.
What emerges is a new understanding of climate-evolution dynamics. Evolution is not a continuous process that simply accelerates or decelerates under environmental pressure. It is a threshold-dependent system that operates in two distinct modes. Above the critical population size, natural selection functions as an adaptive engine, filtering genetic variation to match organisms to their environment. Below that size, the system switches to genetic drift, where change becomes random and adaptation impossible.
Climate models now need a new parameter. Predicting which species will survive climate change requires more than projecting temperature increases and habitat shifts. It requires identifying the thresholds where populations cross from the selection regime into the drift regime. The warmest environments in the experiment showed both outcomes, survival and extinction, in different populations. The difference wasn't the heat itself. It was whether populations maintained the critical mass needed to keep evolution functioning as a solution rather than watching it transform into noise.
The research reveals that evolution's power to rescue species from environmental change is real but conditional. It works, and it works quickly, in three to five years rather than millennia. But only when populations stay large enough for natural selection to maintain its organizing force against the randomizing pressure of genetic drift. Cross that threshold, and the mechanism that has sustained life through every previous climate shift in Earth's history simply stops. The engine doesn't stall. It becomes a roulette wheel, spinning genetic changes that lead nowhere but extinction.