Evolution Lab
Evolution Lab
Evolution is not only something that happened long ago. It is happening right now: in hospitals, in forests, and in every child that is born. Run six experiments yourself, then see what happened when scientists ran the real ones.
1 · Bacteria vs. antibiotics
Bacteria vs. antibiotics
Bacteria spread inward from both edges of a plate. Each band toward the centre holds ten times more antibiotic than the one before. A bacterium cannot grow in a band it is not resistant to. But every time a bacterium divides there is a small chance of a mutation, and some mutations make its offspring a little more resistant.
- Generations
- 0
- Highest band reached
- no drug
- Resistance levels
Press play. Each colour is a level of resistance. Watch where the front stalls, and where a single mutant breaks through.
The real experiment
In 2016, a Harvard team poured a giant plate, 120 by 60 centimetres, with bands of the antibiotic trimethoprim rising tenfold toward the centre, up to 1,000 times the dose that normally stops E. coli. Bacteria released at the edges evolved their way into the centre band in less than two weeks, and a camera filmed every step. One surprise: the most resistant mutants did not always win. They were sometimes trapped behind slower-growing, less resistant neighbours.
Why it matters: overusing antibiotics runs this experiment on a global scale. Every use favours the bacteria that can survive it.
The longest evolution experiment
On 24 February 1988, Richard Lenski started 12 identical flasks of E. coli. Every day a little of each culture is moved to fresh food, which makes about 6.6 generations a day. In August 2024 the populations passed 80,000 generations, more than six times as many generations as our species has lived. Around generation 31,000, one population evolved something new: the ability to feed on citrate, which E. coli normally cannot use when oxygen is present.
Why so fast? In good conditions, E. coli can divide about every 20 minutes.
Studies:Baym et al. 2016, Science ↗Blount et al. 2008, PNAS ↗Good et al. 2017, Nature ↗E. coli long-term evolution experiment ↗Antimicrobial resistance ↗
2 · Be the bird
Be the bird
Peppered moths rest on tree bark during the day. Most are pale and speckled; a few carry a mutation that makes them black. You are a hungry bird: click on the moths you can spot. The survivors lay eggs, and their young look like their parents.
The real experiment
This really happened. Black peppered moths were first recorded in Manchester in 1848. As coal soot darkened the trees, they spread fast, and by 1895 about 98% of the moths around Manchester were black. After clean-air laws in the 1950s, the bark grew pale again and black moths became rare. In 2016, geneticists found the mutation itself: a piece of “jumping DNA” that had landed in a gene called cortex, most likely around 1819.
But was it really the birds?
To settle the question, the biologist Michael Majerus released 4,864 moths over six years and watched which ones birds ate. On today’s clean bark, birds took more of the dark moths, enough to explain why black moths have been disappearing.
Studies:van't Hof et al. 2016, Nature ↗Cook et al. 2012, Biology Letters ↗Peppered moth evolution ↗Natural selection ↗
3 · Gene mixer
Gene mixer
Every child gets half of its genes from each parent, but which half is decided by chance. Meet two imaginary creatures, each carrying two copies of five genes. A capital letter is dominant: one copy is enough for the trait to show. Click a letter to change a parent’s genes.
Parent A
Parent B
Child
You are one of about 70 trillion
Each of your parents had 23 pairs of chromosomes and passed on one chromosome from each pair, chosen by chance. That gives 2²³ = 8,388,608 possible combinations from each parent, and about 70 trillion for one couple, before chromosomes even swap pieces with each other (crossing over), which multiplies the number many times more. On top of that, every child is born with about 70 brand-new mutations that neither parent had.
Gregor Mendel discovered these rules in the 1860s by growing and crossing some 28,000 pea plants in a monastery garden. Nobody understood how important his work was until 1900, sixteen years after his death.
Studies:Jónsson et al. 2017, Nature ↗Mendelian inheritance ↗Chromosomal crossover ↗Gregor Mendel ↗
4 · Darwin’s finches
Darwin’s finches: you control the weather
On Daphne Major, a tiny island in the Galápagos, Peter and Rosemary Grant caught, measured and followed almost every finch for decades. Here you decide each year’s weather. The weather decides which seeds are left, and the seeds decide which beaks survive.
Simplified model: survival depends on how well a beak suits the seeds left that year, and beak depth is passed on with a heritability of 0.75, close to what the Grants measured. The replay follows the real sequence of key years, but the numbers are simulated.
The real experiment
In 1977 only about 24 millimetres of rain fell on Daphne Major. The small, soft seeds ran out, and most of the island’s medium ground finches died. The survivors were mostly large birds with deep beaks, able to crack the big, hard seeds of the caltrop plant. Their offspring had beaks about 4% deeper than the population before the drought: evolution measured with calipers, within a single generation.
Then evolution changed direction
The El Niño rains of 1982–83 flooded the island with small seeds, and in the following years smaller beaks did better. In 1982 a few large ground finches had settled on the island. In the 2004 drought they took the big seeds, and the medium ground finches’ beaks became smaller: the strongest evolutionary change the Grants recorded in 33 years. In 2016 geneticists linked this change to a gene region called HMGA2. The Grants’ lesson after 30 years: in the short term evolution can be predicted; in the long term it cannot, because the weather cannot.
Studies:Boag & Grant 1981, Science ↗Grant & Grant 2002, Science ↗Grant & Grant 2006, Science ↗Lamichhaney et al. 2016, Science ↗Darwin's finches ↗
5 · Tame a fox
Breed a tame fox
In 1959 the Soviet geneticist Dmitry Belyaev began an experiment in Siberia. Of the silver foxes bred on fur farms, only the calmest and friendliest were allowed to have pups. Now you are the breeder: choose the parents of each new generation, and see what else changes along the way.
Simplified model: tameness depends on many genes plus chance. White patches, floppy ears and curly tails appear rarely, and a little more often in tame foxes, as was seen in the experiment; why this happens is still debated.
The real experiment
In the real experiment, fewer than 10% of each generation became parents. In the fourth generation the first pups wagged their tails at people. The “elite”, foxes that seek out humans, whine and lick like dogs, made up 1.8% of pups in generation 6, 18% in generation 10, 35% in generation 20 and 49% in generation 30. By 2005–2006 almost the whole population was elite, and their stress-hormone levels had fallen to a third of those of ordinary farm foxes, or less.
Floppy ears that nobody chose
Only behaviour was selected, yet after 8–10 generations white “star” patches appeared, and some foxes developed floppy ears and curly tails, traits found in many domestic animals. Belyaev believed that selecting for tameness alone changes how genes are switched on during development. That idea is now debated: in 2020 researchers traced Belyaev’s foxes to fur farms in eastern Canada, where foxes had already been bred for decades, and argued that the experiment’s conclusions have been overstated. In 2018 the fox genome revealed regions linked to tame and aggressive behaviour.
Studies:Trut et al. 2009, BioEssays ↗Kukekova et al. 2018, Nature Ecology & Evolution ↗Lord et al. 2020, Trends in Ecology & Evolution ↗Domesticated silver fox ↗
6 · Your own world
Build your own world
Small imaginary creatures live here. They eat, flee, grow old and have offspring. Each carries four genes, for body size, speed, colour and fur, and passes them on with tiny random changes. Nobody designs them. Change their world, and watch natural selection reshape them.
Try this
- Set the ground to pale sand and add many predators. Wait a while, then switch to dark lava. What happens to the creatures’ colour?
- Start an ice age. Who survives, the thick-furred or the thin-furred? Then switch to desert heat.
- Switch food from plentiful to scarce and watch the population chart. How many creatures can the world feed?
- Press “Start again” and just watch. The first creatures are medium-sized and medium-fast. Where do size and speed settle on their own, and why?
This is a simulation, not a real ecosystem, but it follows the same three rules as life: variation, inheritance, and unequal survival and reproduction.
Every gene has a price
A bigger body is harder for a predator to overpower but needs more food. Speed helps to escape and to find food, but burns energy. Fur protects against cold but overheats in the desert. Camouflage only works on the right ground. Selection never finds a perfect creature; it finds a compromise that fits the world as it is now. Change the world, and the compromise changes.
Nothing here is programmed to evolve
The code contains no instruction like “become darker” or “grow fur”. It only says how creatures move, eat, pay for their bodies, get caught and reproduce, with small copying errors. Every trend in the charts emerges from that alone, which is exactly Darwin’s point.
Learn more:Natural selection ↗Camouflage ↗