Science
How coffee is harvested
Why cherries on one branch ripen at different times, the four ways coffee is picked, and what slope and the cost of labour decide.
In short
The difficulty of harvesting coffee starts on the branch: a tree that flowered several times carries green, ripe and dried fruit together. Every way of picking is an answer to that, and which answer a farm uses is decided mostly by its slope and by what labour costs, not by preference.
No method is better in general. Each gives up something, and the sources read here measure time and labour much more often than they measure what ends up in the cup.
Why one branch holds fruit of three ages
Sourced
A coffee tree does not flower once. A review of coffee physiology describes two to four or more flowerings in a season in south-central Brazil, and says ripening can spread over about ten weeks. Where rain falls all year the flowering, and so the harvest, is scattered further. The review names two consequences of a tree carrying green, ripe and dry fruit at harvest: more than one picking is needed, which raises the cost, and a harvest with a low share of ripe cherries gives, in its words, an impaired beverage quality and more defective beans.
A count from one Brazilian farm shows what that looks like at harvest: about a fifth of the fruit was green, under a third ripe and nearly half already dried on the tree, and more than a quarter of the crop had fallen to the ground after unseasonal rain. That is one farm in one year, reported in a study of harvesting methods, but it is the problem every method has to deal with.
Four ways of taking the fruit off the tree
Sourced
The names and descriptions follow the atlas entry for Brazil and a 2019 study from Minas Gerais.
- Selective picking by hand
- Only ripe cherries are taken, and the pickers return as more ripen, so it needs repeated passes. It is how the Colombia entry describes that country's whole crop being picked; in the Brazilian survey described below, a tenth of the growers asked said they used it.
- Stripping by hand
- Everything on the branch is taken in one pass, ripe or not. In the Brazilian study the fruit was stripped onto a cloth laid under the tree.
- Hand-held machines
- A motorised tool carried by the picker shakes the fruit off the branch. It does the same job as stripping and does not need ground a machine can drive on.
- Harvesting machines
- A machine straddles or runs beside the row and shakes the fruit off. It needs ground level enough to drive on.
The ground decides more than the grower does
Sourced
The study from Minas Gerais gives the working limit for ordinary harvesters as a slope of 20 per cent. Above that, growers in the region have cut narrow terraces between the rows so that tractors and one-sided harvesters can pass, and the study was made to find out whether that works. It compared a one-sided harvester built for terraces with two experienced pickers stripping by hand and with a hand-held machine, on one terraced farm with a slope of 47 per cent. By its authors' reckoning the harvester did the work of about 24 hand pickers, or of about 11 people with hand-held machines. It also stripped about half as much leaf from the trees as hand stripping did, a difference the study found significant; the hand-held machine fell in between and could not be told apart from either. The harvester dropped some fruit on the ground, which then had to be swept up. The authors conclude that mechanising is preferable even on slopes, and they give the shortage of labour in mountain areas as the reason growers want it.
What the study did not measure is the ripeness of what each method brought in, the quality of the coffee, or money. Its conclusion is about working time, capacity and damage to the tree, on one farm in one season.
A survey of 296 Brazilian growers reported in 2019, described in the Brazil entry, found all four methods in ordinary use. It counted farmers, not coffee, so it does not say how much of the crop each method brings in.
What selective picking costs
Sourced
Where coffee is picked ripe by hand, picking is the largest cost of growing it. The Colombia entry reports the growers' federation's figures: picking is 52 per cent of the cost of production and labour of all kinds 69 per cent. It also reports why the passes have to be frequent there: the main varieties descend from Caturra, which sheds its ripe fruit easily.
That makes the supply of pickers a grower's main exposure. The entries for Colombia and Nicaragua both describe harvests short of people, as their sources report them.
What each method asks for and gives up
CoffeeHQ explanation
A model for comparing the methods, assembled from the sources above. It is not a ranking and has no measurements in it.
| Variable | Typical range | What moving it does |
|---|---|---|
| Selective picking | Many people, repeated passes | Takes ripe fruit only and needs no machine. The most labour of any method. |
| Stripping by hand | Fewer people, one pass | Takes fruit of every ripeness together, so sorting has to happen later or not at all. |
| Hand-held machines | Fewer people still, one pass | No vehicle on the ground. Same mixed fruit as stripping. |
| Harvesting machines | Very few people, one or more passes | Needs level or terraced ground. Takes whatever the machine shakes loose. |
A model for explanation. No source read measured the cup quality of coffee harvested by each method side by side, so the table says nothing about which tastes better.
What is left on the tree matters too
Sourced
Dried fruit left on the tree or the ground after harvest can harbour the coffee berry borer. A review of the pest lists clearing those leftovers among the measures used against it in the countries it describes. A harvest is therefore also a piece of pest control, and what a method leaves behind is part of its cost.
What nobody measured: the cup
CoffeeHQ explanation
It does not say that hand-picked coffee tastes better, because no study comparing the methods on quality was read; what it reports is the physiology review's statement that a harvest short of ripe fruit makes worse coffee, which is about what is picked and not about who or what picks it. It does not describe how ripeness is judged, how cherries are sorted after picking, what pickers are paid, or the share of the world's coffee harvested by each method. Harvest months for each country are in the atlas, each with its source.
What to do next
- Harvest timing in the atlas — The hub's harvest view lists each country's sourced harvest months side by side.
- Brazil in the atlas — Where stripping and machines are the norm, and why.
- Colombia in the atlas — Where every cherry is picked by hand and picking is half the cost.
- Processing — What happens to the fruit in the hours after it is picked.
- Coffee pests and diseases — The berry borer, and why what is left on the tree matters.