The Genome Remembers Place

Marine three-spined sticklebacks were trapped in some of these new lakes. They were ocean fish. They had spent millions of years optimizing for saltwater — the physics of osmosis, the particular predators, the food chains, the armor. Bony plates running the length of their bodies, useful in the ocean against predators that grab and bite, costly in freshwater where the metabolic overhead outweighs the threat. And then the ground moved.

Within 20 to 30 years, the lake populations had changed. The bony armor started to disappear. Internal architecture shifted. By any functional measure, these looked like freshwater fish now.

Except: the genes for freshwater adaptation were already in the marine population, dormant, present as standing variation across approximately 100 genomic regions. The earthquake didn't force evolution from scratch. It woke up a solution the genome had been holding in reserve for millions of years — a contingency plan the fish never needed until one Friday in March.

The genome, it turns out, remembers places it has been. Not as metaphor. As mechanism.

i · the part of the deck that won't shuffle

The word for what makes this possible is "inversion," and it is worth understanding precisely. When a segment of DNA breaks off a chromosome, rotates 180 degrees, and reinserts itself, it produces an inversion. The genes are still present, the sequence is intact — just running backward relative to the rest of the chromosome.

The consequence of this small structural event is enormous. During reproduction, homologous chromosomes pair up and swap segments in a process called recombination — this is what shuffles the genetic deck, the mechanism that creates new combinations of traits in offspring. Recombination is the engine of variation. But an inverted segment cannot properly pair with its non-inverted counterpart. The chromosomes recognize the mismatch. The swap doesn't happen. That chunk of DNA refuses to participate in the shuffle.

What you get is a "supergene" — a cluster of different genes permanently locked together, inherited as an indivisible unit, never broken apart. If those genes happen to work together to solve an environmental problem — saltwater versus freshwater, wave-battered coast versus sheltered cove, open meadow versus dense forest canopy — then natural selection can act on the whole package at once. A complete environmental solution, sealed in molecular amber, preserved across geological time.

"You have a part of the deck that refuses to shuffle," explained Patrik Nosil, an evolutionary biologist studying these structures, "so you can never change the order of the cards."

This is the architecture behind what biologists call ecotypes — distinct populations of the same species, adapted to specific local environments, carrying their adaptations as locked genomic units rather than diffuse mutations scattered across the chromosomes. And as whole-genome sequencing has made it possible to read these structures precisely, a pattern has emerged that is genuinely strange: the same ancient inversions appear in populations separated by thousands of miles and millions of years, as if the genome had been running the same contingency plan across continents without anyone coordinating the effort.

ii · the snails who remembered a coastline

Botanist Göte Turesson noticed in 1922 that Swedish saltbush plants growing in different habitats had different traits — and that these differences persisted when the plants were grown in identical laboratory conditions. The traits weren't just responses to immediate environment. They were carried genetically. He called these habitat-specific subtypes "ecotypes," and the term stuck even though the molecular tools to explain the mechanism were six decades away.

Kerstin Johannesson has spent decades studying marine snails — specifically Littorina saxatilis, the rough periwinkle, which presents in two recognizable ecotypes along European coastlines. On wave-battered rocky shores: smaller, thin-shelled, adapted to cling in crevices and escape crab predation through mobility. In sheltered coves: larger, thick-shelled, built for direct resistance. The two ecotypes are the same species. They can interbreed. But in their respective environments, the differences are sharp enough to look like what a textbook would call speciation.

In 1988, an algal bloom disrupted the local predator community in one of Johannesson's study areas. She relocated thick-shelled snails to wave-exposed rocks. Within 30 years, the population had shifted toward the thin-shelled ecotype — not because the snails themselves changed, but because the thin-shelled genetic variation was already present in the thick-shelled population as latent standing variation. Selection pressure revealed what was already there.

When whole-genome sequencing became available and researchers finally read the molecular architecture beneath these differences, they found approximately 20 inversions associated with the ecotype variation in Littorina saxatilis. Already notable. But more striking: the same inversions appear in snail populations across Spain, the United Kingdom, and Sweden — populations geographically separated for at least a million years. The coastline kept presenting the same problem: wave-battered rock versus sheltered cove. The genome kept producing the same answer, not by reinventing it each time, but by reaching back to a solution encoded before the populations diverged.

The genome was running a distributed backup. Every separated population carried the full contingency archive, waiting for the right environment to make either file active.

iii · what the earthquake released

Return to the sticklebacks and March 1964.

The 50-year genomic study that followed those earthquake-created lake populations found standing variation across roughly 100 genomic regions associated with freshwater traits in the founding marine population. The genes for reducing bony armor. The genes for freshwater osmotic regulation. All present, all maintained through generations of marine existence, never fully purged.

This is not neutral persistence. Genes that carry no benefit and some cost get eliminated by selection over evolutionary time. If these freshwater genes stuck around in the marine populations, it's because occasionally — in estuaries, in river mouths, in seasonal pools — having them mattered just enough. The genome held the solution because the problem, though rare, was not extinct. And then 1964 created a lot of freshwater lakes very quickly, and the genome was ready.

Catherine Peichel at the University of Bern is now using CRISPR to reverse specific inversions in sticklebacks, watching what happens to phenotype when the supergene architecture is undone. The goal is mechanistic confirmation: that the deck-that-won't-shuffle is actually the structure doing the remembering, not just a correlate of something else. The work is ongoing. The hypothesis is not fragile.

Sean Stankowski at University College London summarized the moment plainly: "It's by far the most exciting time to be a biologist, ever."

That is not modesty. Whole-genome sequencing arrived like a new sense organ — suddenly biologists could read the structural history of populations across millions of years, find the same inversions in populations separated by continents and geological epochs, and start building a picture of how evolution actually stores and retrieves its solutions. Not just random mutation and selection grinding forward, but something more like a library organized by environment, with a retrieval system that activates when the right conditions reappear.

iv · what darwin's finches actually are

The Galápagos finches are the iconic speciation story: isolated populations adapt to different food sources, beaks diverge, distinct species emerge. That's the narrative most of us carry. The ecotype research complicates it.

Darwin's finches exhibit rapid morphological change, maintain gene flow between populations, and can interbreed across apparent divergences. Peter and Rosemary Grant spent 40 years on Daphne Major Island documenting exactly this — recognized Geospiza species hybridizing during climate extremes, hybrid lineages persisting and sometimes outcompeting parental forms. That's not the signature of fully separated species — that's what you'd expect from ecotypes living on different islands. Cichlid fish, which diversified into hundreds of apparent species in African rift lakes in what looks like evolutionary blink speed, show similar patterns. Rapid change. Maintained interbreeding capacity. Standing genetic variation expressed differently in different local environments.

The boundary between species, subspecies, and ecotype is — and researchers working in this space are fairly direct about this — more culturally determined within scientific communities than biologically discrete. Different fields have different conventions for when to call something a new species. The biology doesn't sort itself neatly into the categories.

What the genome reveals beneath those categories is something more interesting than tidy tree diagrams allow: life doesn't always wait for slow mutational grinding to generate new solutions to new problems. It stores solutions, in locked inversions, in maintained standing variation, in genomic architecture that persists across millions of years — and when the environment presents the problem again, it reaches back for what worked. The genome is not just a blueprint for the current organism. It is a record of everywhere the lineage has been, organized by what was useful there.

The 1964 earthquake made new lakes. The fish already knew what lakes required. The ground moved. The archive opened.

v · sources

• How Ecotypes Harbor the Genetic Memory of a Species' Past — Quanta Magazine, 2026-05-21
• Littorina saxatilis — Wikipedia
• Three-spined stickleback — Wikipedia