When California's most severe drought in 10,000 years descended in 2012, the scarlet monkeyflower (*Mimulus cardinalis*) should have been doomed. A potted specimen without water dies within days. Yet multiple natural populations along California's river corridors survived — and researchers at **Cornell University** and the **University of British Columbia** have now discovered exactly why.
The answer: the flowers evolved. Fast. Fast enough to rescue their own populations from extinction.
The findings, published March 12 in *Science*, represent the **first full documentation of "evolutionary rescue" from climate change in natural wild populations** — a phenomenon previously confirmed only in labs and mathematical models. The implications for conservation biology are extraordinary.
**A Ten-Year Natural Experiment**
The story begins in 2010, when senior author **Amy Angert** (UBC) and then-PhD student **Seema Sheth** began tracking variations in scarlet monkeyflower populations across 55 sites in Oregon and California — with no expectation that history was about to intervene.
Then came the drought: a four-year, record-shattering dry spell that began in 2012, the worst in over a millennium.
"We had a time capsule in the form of stored seeds back in the lab," said first author **Daniel Anstett**, now assistant professor of plant biology at Cornell. As populations crashed, the team was able to compare pre-drought and during-drought genomes across all 55 sites, using whole-genome sequencing to track which genetic variants were present and which populations survived.
**What They Found**
The results were striking on two levels:
1. **Populations that survived are also the ones that evolved fastest.** In the populations that recovered, scientists found rapid shifts in climate-associated gene variants — the exact genetic changes that confer drought tolerance. The populations that didn't evolve fast enough died.
2. **The genetic variation *before* the drought predicted which populations would survive — years later.** This is the result that stunned the team. Genetic snapshots taken in 2010, two years before the drought began, contained signals that predicted which populations would eventually recover — five, six, even seven years into the future.
"The genetic variation we saw, even before the drought, predicted demographic recovery five, six, seven years later," Anstett said. **"That's astounding. That's the crystal ball we can use to predict into the future."**
**Evolutionary Rescue: What It Is**
Evolutionary rescue is the process by which a population avoids extinction not by luck or migration, but by rapid genetic adaptation. The idea has been theorised for decades, but actually documenting it in wild populations — across whole genomes, in real time, under climate pressure — has never been done before.
Not all populations made it. Some went extinct. Evolutionary rescue is real, but it is not universal. The critical factor was whether populations had enough pre-existing genetic diversity — "standing variation" — to draw on when conditions changed.
**The Conservation Implication**
Conservationists working with endangered species can now potentially use genomic data to predict which populations are likely to survive climate extremes — and which urgently need intervention.
"Conservation is a complicated calculus, and this gives greater information to decision-makers," Anstett said. "It's another arrow in our quiver to try to conserve species."
The monkeyflower is one species, in one region, under one kind of climate pressure. But the methodology — genome sequencing + long-term population tracking + demographic data — is transferable to other species, other stresses, other habitats.
At a time when climate change feels overwhelming and species loss feels unstoppable, this study is a reminder that nature has its own toolkit. Life doesn't just endure — sometimes it adapts faster than we thought possible. 🌸🧬
*Sources: Cornell University · Science (journal, DOI: 10.1126/science.adu0995) · University of British Columbia · North Carolina State University · March 2026*
