# From Flower to Harvest (/en/docs/foundation/phase-1/chapter-06)
## Phenomenon [#phenomenon]
One morning in harvest season, a picker stands in front of a coffee plant. The same branch carries deep-red cherries, bright-red ones, yellow-green ones and a few small, hard green ones. She takes only the red ones and leaves the rest on the branch — in a week or two she will come back, and then come back again.
On a coffee farm in another country, a machine straddles an entire row of plants and shakes the branches. Cherries rain down into a collector: red, green and half-dried, all mixed together.
Why does one farm pick the same plant several times, while the other picks it all at once? If harvest is just getting fruit off a tree, what makes it hard?
## Variables [#variables]
From flowering to harvest, the variables that shape "what gets picked":
| Variable | What it decides |
| ---------------------------------- | ---------------------------------------------------------------------------------------------- |
| How concentrated the flowering was | Whether the fruit ripens at the same time (Chapter 2) |
| Stage of ripeness | Green → yellow → red → deep red → starting to shrivel — which moment it is picked at |
| Harvest method | Selective hand-picking, strip-picking (stripping a whole branch in one go), mechanical harvest |
| Number of passes | How many times the same plant is picked |
| Labor and cost | Whether pickers can be hired, and how they are paid |
| Terrain | Whether a machine can get in |
| Post-harvest sorting | Flotation, hand sorting, color sorting — how much unripe fruit can be removed |
## Experiment [#experiment]
Thought experiment one: on one branch, 30% of the cherries are still green, 50% are just ripe and 20% are overripe and starting to shrivel.
Plan A: wait until most are ripe, then strip the whole branch in one go. Plan B: come back at intervals and take only the ripe ones each time, finishing in three passes. How much of each harvest will be ripe? Which needs more labor? Which is likely to make better coffee?
Thought experiment two: from the same plant, take one green cherry and one fully ripe red one, and process, dry and roast them the same way.
After roasting, will the two seeds be the same color? Which will weigh more? Ground and brewed, how might they taste different?
## Observation [#observation]
**Ripening is a process, and color changes last.** Early in fruit development, the seed's endosperm is liquid; as the fruit matures, it accumulates storage proteins, sucrose and polysaccharides and gradually hardens; only at the very end does the endosperm dry down and the skin change color. So the seed inside a green cherry has not finished filling.
**Color is not always reliable either.** In a study in Costa Rica, fruit flesh in full sun ripened earlier than under shade, and those seeds showed signs of incomplete maturation, giving a more bitter, astringent cup. How fast the skin turns red and how fast the seed fills do not always line up.
**Unripe cherries make defective beans.** Research has found that green coffee from unripe cherries is lighter, holds less moisture and contains less sucrose than normal beans, and lowers cup quality. The trade calls beans that stay noticeably pale after roasting and will not take color "quakers"; they mostly come from unripe cherries, and they often taste of peanut, cereal or cardboard.
**Overripe is a problem too.** Cherries that have gone past ripe and started shriveling on the plant may already be fermenting or spoiling, which can bring over-fermented flavors.
So the answer to the first experiment: Plan B gives a far higher share of ripe fruit, at the price of several times the labor; Plan A saves labor but carries green and overripe fruit into the mill along with the ripe.
## Model [#model]
### Ripeness is not a moment but a distribution [#ripeness-is-not-a-moment-but-a-distribution]
Think of a plant as a **ripeness distribution**: at any moment, some of its fruit is unripe, some just right, some past it. How "wide" that distribution is depends mainly on how spread out the flowering was (Chapter 2), and flowering depends on the rain (Chapter 4).
Harvest is a set of choices made on that distribution: **on which day, with which method, to take which part of it.** That is why ripeness is the hard problem — not because cherries are hard to pick, but because they do not ripen together.
### Three methods, three trade-offs [#three-methods-three-trade-offs]
Coffee is harvested in roughly three ways:
| Method | How it works | Ripeness control | Cost |
| ---------------------- | -------------------------------------------------------------------- | ------------------------------ | ----------------------------------------------------- |
| Selective hand-picking | Only ripe cherries each time; the same plant is picked several times | Best | Labor-intensive, expensive, needs skilled pickers |
| Strip-picking | Wait until most are ripe, then strip each branch in one go | Poorer — ripe and unripe mixed | Less labor, but relies on sorting afterwards |
| Mechanical harvest | A machine shakes the branches so the fruit falls | Mixed, but adjustable | Needs gentle terrain and large areas; high investment |
Ripe cherries come off the branch more easily than unripe ones, so tuning how the machine shakes can make it favor ripe fruit, and the harvest can be split into several passes. Mechanical harvest does not have to mean careless harvest; it selects less precisely than a skilled hand, but its efficiency and cost are in a different league.
### How much can be fixed after picking? [#how-much-can-be-fixed-after-picking]
Picked cherries are often sorted further: in water, the floaters tend to be unripe, insect-damaged or half-dried fruit; then hand sorting or optical color sorters remove cherries or beans of the wrong color.
Sorting can only remove unripe fruit; it cannot make it ripe. The stricter the sorting, the more weight is lost; the looser, the more defects slip through.
### Why this is so hard [#why-this-is-so-hard]
In the end, the ripeness problem comes down to people and money:
* Pickers are often paid by the weight they pick, and that encourages picking everything, not just the ripe fruit;
* Every extra pass means another round of wages;
* The weather does not wait: if the rains come early, fruit may not be picked in time;
* Some producers measure the sugar in cherry juice with a refractometer as a guide to ripeness, but color, feel and experience are still the main tools.
So when a bag says "hand-picked ripe cherries", part of its price is paying for those extra rounds of labor.
## Transfer [#transfer]
**Follow one cherry.** This is the Phase I project. From Chapter 2 to this one, you can now tell the life story of a single coffee cherry:
1. A rain after the dry season wakes the flower buds on one branch (Chapters 2 and 4);
2. The flower opens and is pollinated, and the fruit begins to grow; the seed starts out liquid, then fills and hardens (Chapter 2);
3. The plot's temperature, slope and shade decide how fast it grows (Chapters 4 and 5);
4. Its species and variety decide what it can become (Chapter 3);
5. Most of a year later, it is picked on one particular day — how ripe was it at that moment?
Read the bag the same way:
| On the bag | What it tells you | What it cannot promise |
| --------------------------------------- | ---------------------------------- | -------------------------------------------------------- |
| Hand-picked | Someone picked the fruit by hand | Only ripe fruit — hand-picking can also be strip-picking |
| Selectively picked / ripe cherries only | Ripeness was controlled at harvest | That every single one was fully ripe |
| Mechanically harvested | It was picked by machine | "Worse" — it depends on tuning and sorting |
| Harvest season | When the fruit was picked | How fresh the green coffee is now (Chapter 12) |
## Misconceptions [#misconceptions]
**"Once the cherry turns red, the seed inside is fully ripe."** Color is the last step of ripening and the most convenient signal, but not a perfect one. Under conditions such as full sun, the skin can "ripen" ahead of the seed.
**"Hand-picked is always better than machine-harvested."** Hand-picking allows the best ripeness control — but only if pickers take ripe fruit alone, over several passes. Strip-picking is also done by hand, yet takes ripe and unripe together. Mechanical harvest can be tuned and combined with sorting, and on large, gently sloping farms it is a reasonable choice.
**"Harvest is just getting the fruit off the tree; it has little to do with flavor."** The moment of picking fixes how much the seed has stored; neither processing nor roasting can add it back. Unripe cherries become defective beans, and overripe ones can bring over-fermented flavors.
## Real Lab [#real-lab]
There is no registered lab card for this chapter, but you can find traces of the harvest in a bag of roasted beans.
1. Pour out about 100 g of roasted beans onto white paper, in good light.
2. Pick out the beans that are clearly paler than the rest — yellowish or whitish. These may be quakers from unripe cherries. Count them.
3. Compare a few different bags (they are harder to spot in light roasts, easier in medium roasts). Which bag has the most quakers? Compare with the harvest method and price on each bag — what do you notice?
4. (Optional) Snap a quaker open and smell it, or collect a small pinch, grind it and steep it in hot water, and taste it against normal beans.
If you ever see a coffee plant at origin, notice how many colors of fruit hang on one branch, and ask yourself: if today were harvest day, which ones would you pick?
## Check Your Model [#check-your-model]
**Predict.** To save money, a farm switches from "three passes of hand-picking, ripe fruit only" to "wait until most are ripe, then strip everything at once". What changes do you expect in the green coffee and the cup?
More green and overripe cherries will end up in the harvest. Unless sorting is very strict, the green coffee will contain more unripe beans; after roasting you may see more pale quakers, and the cup is more likely to taste green, cereal-like or over-fermented. With very strict sorting, quality can be partly protected, but more weight is lost. The labor saved has to be weighed against the loss of quality and weight.
**Diagnose.** A bag you bought has quite a few pale beans, and the brewed coffee tastes of peanuts and cardboard. Where on the chain is the most likely cause?
Most likely at harvest: pale beans that will not take color often come from unripe cherries whose seeds had not finished filling, so they cannot brown and develop aroma normally in the roaster. Sorting after harvest may also have been too loose to remove them. This is not a brewing problem — changing grind or water temperature will not remove that taste.
**Explain.** In your own words, explain why ripeness, rather than picking itself, is the real problem of harvest.
Getting fruit off a plant is not technically hard. What is hard is that the cherries on one plant, even on one branch, do not ripen together, because the flowers were woken in waves by successive rains. Every harvest method is a trade-off on that ripeness distribution: picking only ripe fruit means several passes and high cost; picking everything at once saves labor but takes unripe and overripe fruit too. The moment of picking fixes what the seed has stored, and nothing downstream can add it back.