Last lesson handed you a slogan: variation, selection, heredity — repeat. A slogan is a fine doorway and a useless tool. To actually use natural selection — to spot it in a market, predict a population, or debug why some “design” keeps appearing — you need to know the machine part by part: what each ingredient does, and what happens to the whole contraption when one of them goes missing. That last bit is the real test. Anyone can recite three words; the person who understands the recipe is the one who can tell you why removing any single one of them brings the entire process to a dead stop.
So that’s this lesson’s job. We’ll take the recipe apart, prove why each ingredient is load-bearing, and then watch a trait actually climb through a population with numbers you can add up yourself. As always, commit to a guess before you peek.
Before you read — take a guess
Natural selection needs three ingredients: variation, selection, and heredity. Suppose a population had selection and heredity but NO variation — every member was a perfect identical copy. What would happen over the generations?
Ingredient 1 — Variation
The analogy. Think of variation as the menu selection gets to order from. A restaurant with one dish on the menu offers no real choice — you eat that dish or you starve. A restaurant with fifty dishes lets a discerning diner pick the best one. Variation is the length of the menu: the more the members of a population differ from one another, the more raw material selection has to work with.
The precise definition. Variation means the members of a population are not identical — they differ in their traits (size, colour, speed, shape, behaviour, biochemistry). Crucially, some of that difference must be heritable (more on that in ingredient 3), because variation that can’t be passed on is a dead end. Where does variation come from? In biology, mainly from mutation (small random copying errors in DNA) and the genetic reshuffling of sexual reproduction. You don’t need the molecular details — just the fact that real populations are full of differences, and those differences arrive blindly, with no eye on what would be useful.
Why it’s necessary. Selection can only choose among options that exist. No variation, no options, no change — exactly the pretest. This is why a population of perfect clones is an evolutionary dead end: there’s nothing for the filter to grab.
Variation is blind — it doesn't aim
A subtle but crucial point we’ll hammer in lesson 5: variation does not arrive because it’s needed. Mutations don’t show up in order to solve a problem — they happen randomly, most are neutral or harmful, and only occasionally is one useful. The beetle didn’t mutate darker because birds were hunting; dark variants were already kicking around the population by chance, and then the birds did the choosing. Variation proposes blindly; selection disposes. Keep those jobs separate and half the misconceptions about evolution evaporate.
A bacteria colony is hit with an antibiotic. A few cells happen to carry a mutation that resists the drug; they survive and repopulate. Which statement matches how variation actually works?
Ingredient 2 — Selection
The analogy. If variation is the menu, selection is the diner with preferences — except the “preference” here isn’t a choice anyone makes. It’s simply that some variants survive and reproduce more than others in a given environment. The environment is the diner, and “being eaten less” or “finding more food” or “attracting more mates” is what gets a variant ordered again next generation.
The precise definition. Selection means the trait differences cause differences in survival and reproduction. A variant that, in this environment, leaves more surviving offspring than its rivals will make up a larger slice of the next generation. That’s the entire mechanism — there’s no judge, no scorekeeper, no intention. It’s just the bookkeeping of who managed to copy themselves more.
Why it’s necessary. With variation and heredity but no selection — if every variant reproduced equally well — the proportions would just wobble around at random (this is called genetic drift, and it’s real, but it’s aimless). Selection is the ingredient that makes change directional: it’s the difference between a population marching toward a trait and one wandering in place.
You can feel this directly. In the simulator below, set the selection pressure to the middle (no selection) and hit advance a few times: the average shade just jitters around, going nowhere. Now slide the pressure hard toward “favours dark” and advance again: suddenly the population marches. Same variation, same heredity — selection is the only thing you changed, and it converted aimless drift into directed design.
Knock out an ingredient
Selection is the difference between marching and drifting
Start with the pressure in the middle — no selection. Advance a few generations and watch the average just wander. Then drag the pressure to one side and advance again: now it marches. Selection is what turns variation into direction.
Generation 0: average shade is 49/100, trending nowhere — just drifting.
Two islands have beetles that vary in colour and pass colour to their offspring. On Island A, colour makes no difference to survival — every beetle breeds equally. On Island B, dark beetles survive better. What's the key difference after many generations?
When to use it
Whenever you see a population (of organisms, products, ideas, anything) that has clearly moved in a consistent direction over time, ask: what was the selection pressure? Something must have made one variant reproduce or persist better than the others. And conversely, when something is just churning with no consistent trend, suspect there’s no selection pressure — only drift.
Ingredient 3 — Heredity
The analogy. Heredity is the save button. Imagine playing a game where every good move you make is erased the instant you make it — you could never build on progress, because nothing carries forward. Heredity is what saves a useful variation so the next generation starts from it instead of from scratch. Without a save button, every generation reinvents the wheel and selection’s work is wiped each round.
The precise definition. Heredity (or inheritance) means offspring tend to resemble their parents — traits are passed down, so a parent who survived because it was dark tends to have dark offspring. In biology the carrier is the gene, but again you don’t need the chemistry: the load-bearing fact is just that traits are transmitted across generations with enough fidelity that selection’s choices stick.
Why it’s necessary. This is the subtlest of the three, so go slow. Suppose you had variation and selection but no heredity — offspring colour was random, unrelated to the parents’. Selection would dutifully kill the pale beetles every generation… and every generation, the survivors would have offspring of totally random colours, so you’d be right back to 50/50. Selection’s hard-won progress would evaporate every single round. Heredity is what makes the gains cumulative — it’s the ratchet that stops the population from sliding back.
Look at what each ingredient supplies, and why no two can cover for a missing third:
- Variation supplies the raw material — the options. (No options → nothing to choose.)
- Selection supplies the direction — which options win. (No selection → aimless drift.)
- Heredity supplies the memory — winners carry forward. (No heredity → progress erased each round.)
Raw material, direction, and memory. Remove any one and the machine stalls in a different way: no material to work with, no direction to move in, or no memory to keep what was gained. That’s why “variation, selection, heredity” isn’t a list to memorize — it’s three gears that only turn together. Repeat is just letting them turn.
Each scenario is missing one ingredient of natural selection. Sort each by which ingredient it lacks — and the process fails in that ingredient's specific way.
Place each item in the right group.
- Successful traders make money, but their kids inherit none of the skill or the strategy
- Beetles vary and inherit colour, but colour has zero effect on who survives or breeds
- Web-page designs differ and get copied, but visitors convert at identical rates on all of them
- A population of genetically identical cloned trees in a plantation
- Fast antelopes survive better, but speed is determined by random luck, not passed to calves
- Every smartphone model is an exact copy of one another with no design differences
Press repeat: cumulative adaptation, in actual numbers
Here’s where the magic stops being hand-wavy. “The population gets darker” is vague; let’s count it. Take 100 beetles, half dark and half pale. Dark beetles hide from birds, so 90% of them survive to breed each generation; pale beetles stand out, so only 60% survive. Survivors reproduce to refill the population to 100, and offspring inherit their parent’s colour. Watch what blind bookkeeping does over just five generations:
| Generation | Dark beetles | Pale beetles | Dark % of population |
|---|---|---|---|
| 0 (start) | 50 | 50 | 50% |
| 1 | 60 | 40 | 60% |
| 2 | 69 | 31 | 69% |
| 3 | 77 | 23 | 77% |
| 4 | 83 | 17 | 83% |
Trace one step so you trust it. In generation 0, the 50 dark survive at 90% (→ 45) and the 50 pale at 60% (→ 30). That’s 75 survivors, of which 45/75 = 60% are dark. They breed back up to 100, offspring inheriting colour, so generation 1 is 60 dark, 40 pale. Run the exact same arithmetic again and dark climbs to 69%, then 77%, then 83%. No beetle changed colour. No beetle tried. A consistent survival edge, compounded across generations by heredity, is the design process — and like compound interest, it starts slow and then runs away. Give it a few hundred generations and “pale” is essentially gone.
Selection is a ratchet, not a wish
The deep takeaway: a small, consistent survival difference, repeated and saved by heredity, accumulates into large directed change — with nobody steering. The beetle population “designed” its own camouflage the way water “designs” a riverbed: not by intending to, but by a filter applied over and over to a varying, heritable population. This is the same compounding engine you met in arithmetic form elsewhere in the latticework — here it compounds adaptations instead of money.
Fill in the recipe and why each part is load-bearing:
Pick the right option for each blank, then check.
Natural selection needs three ingredients. supplies the raw material — without differences, there's nothing to choose between. supplies the direction — without it, traits only drift at random. supplies the memory — without it, each generation's gains are erased and the population resets. The fourth word, , is just letting the three gears turn generation after generation, which is what makes small edges compound into large change.
Recap
You came in with a slogan and you’re leaving with a machine you can take apart:
- Variation — members of a population differ, and some differences are heritable. It’s the raw material; it arrives blindly, not because it’s needed. No variation → nothing to select.
- Selection — some variants survive and reproduce better in a given environment, so they fill more of the next generation. It’s the direction. No selection → aimless drift, not adaptation.
- Heredity — offspring resemble their parents, so winning traits carry forward. It’s the memory / ratchet. No heredity → progress erased every round.
- Repeat — the three gears turning generation after generation. Like compound interest, the power is all in the repetition: small edges become large designs.
Check yourself: the recipe
Why does a population of genetically identical clones never adapt by natural selection, no matter how harsh the environment?
Check your answer to continue.
Where this goes next
You can now name the three ingredients, justify why each is necessary, and count a trait climbing through a population. But we’ve been a little loose with one word — “survive.” The beetles that “win” aren’t the strongest or the fastest or the most impressive; they’re simply the ones that leave the most surviving offspring in this particular environment. That idea has a name — fitness — and it is the single most misunderstood word in all of evolution. Lesson 3 pins it down, and then does something sneaky: it changes the environment and shows how the exact same beetle can go from fittest to doomed without changing at all.