How new cacao trees are made
There are three routes and they produce genuinely different plants. Seed is cheap and every tree is a lottery. Grafting gives you the exact tree you chose, on somebody else's roots. Cuttings give you the exact tree with no roots but its own — and a shape you may not want.
If you read nothing else
Seedlings are variable because cacao does not breed true; grafted and cutting-grown trees are genetically identical to their parent. Which route a farm takes determines what it can predict about the trees it will have in ten years — and predictability is the entire argument.
Why this is a real decision and not a technicality
A cacao tree is a decades-long commitment made once. The choice of how it was propagated decides what the farm can know in advance about yield, disease tolerance and pod characteristics, and it cannot be revisited without starting again.
It is also the point where a replanting programme succeeds or fails. Every intervention aimed at ageing, diseased or unproductive cacao — and there are many, across every producing country — comes down to getting appropriate planting material into the ground. Nothing downstream compensates for getting that wrong.
From seed: cheap, available, and a lottery every time
Cacao is largely outcrossing and does not breed true. A seed from an excellent tree produces a tree that is not that tree — it carries a recombination of two parents, and the pollen parent is usually unknown.
So a seedling field is genuinely variable: some trees bear well, some barely bear, and the difference is not visible until years in. Where seed comes from a controlled cross between known parents that variability is narrowed considerably, which is what a hybrid seed programme is for, and it is still not the uniformity of a clone.
There is a second constraint that shapes everything about how cacao seed is handled. The seed is recalcitrant: it cannot be dried and stored without dying. There is no packet of cacao seed on a shelf and no seed bank for cacao genetics — collections have to be kept as living trees, in fields, exposed to the diseases the collections exist to breed against.
What seedlings do have is a taproot, which grafted and cutting-grown material does not develop in the same way, and which matters in dry periods.
Grafting: the tree you chose, on roots chosen separately
Grafting joins budwood — cut shoot material carrying buds, taken from a specific chosen tree — onto a rootstock seedling. The top grows into a genetic copy of the tree the budwood came from. The roots belong to the seedling.
That separation is the point. The top is selected for what it produces: pod characteristics, tolerance of whatever disease dominates locally, bearing behaviour. The rootstock is selected for vigour and for coping with what is actually in that soil. A grower can therefore plant a clone chosen for its pods on roots chosen for the ground it is going into.
The beans are produced by the grafted top and carry its genetics. Rootstock is sometimes claimed to influence flavour by analogy with wine; in cacao that case is far weaker than the confidence with which it is made, and this catalogue does not assert it.
Grafting is also how an existing tree is converted rather than replaced. Grafting new material onto an established stump — top-working — keeps a mature root system and reaches bearing far sooner than a new planting would. It is one of the main tools in rehabilitating an underperforming block.
Rooted cuttings: the whole plant copied, shape included
A cutting is a piece of branch induced to grow its own roots. The result is genetically identical to its parent with no rootstock involved at all.
The catch is the one from tree architecture: growth habit travels with the material. A cutting taken from a fan branch produces a low, spreading, trunkless plant with a shallow root system, because that is what fan branch tissue does. Cuttings from vertical shoots behave like trees, but vertical shoot material is far less abundant on the trees worth copying.
So cuttings deliver clonal fidelity at the cost of plant form and rooting, which is why grafting is more common where both are available.
The constraint everybody forgets: the trees have to be able to pollinate each other
Many cacao trees are self-incompatible — genetically unable to set fruit from their own pollen, or from the pollen of a genetically identical tree.
For seedlings this is invisible, because every tree in the field is different. For clones it is decisive. A block planted uniformly with a single self-incompatible clone is, from the flowers' point of view, one tree repeated. It can flower magnificently and set almost nothing.
The fix is planting design: compatible clones interplanted so that pollen has somewhere to go. It is not a defect in the clone and not something the grower can correct afterwards, and it is the clearest example in this subject of a biological fact that has to be handled at the layout stage or not at all.
Covered in this guide
- How a cacao tree growsWhy material taken from a fan branch grows into a different plant from material taken from a chupon.
- Planting material and nurseriesWhat decides whether the material being propagated is worth propagating.
- Cacao varieties explainedWhat a clone actually is, and why the familiar three variety names describe trade rather than genetics.
- Cacao swollen shoot virus diseaseThe disease whose only real response is removal and replanting — which makes propagation the whole of the answer.
Sources
- Cocoa — Blackwell Science (Tropical Agriculture Series), 2001(citation identity confirmed; passage not re-read)
- ChocolateHQ editorial synthesis — ChocolateHQ(citation identity confirmed; passage not re-read)