Varieties and Planting Material · Cocoa · Ghana Agribusiness Playbook

Varieties and Planting Material · Pillar 01

Cocoa is a tree crop, so the planting decision is the only one a farmer cannot revise. Ghana’s answer to it has changed twice in a century, and the sales pitch for the current answer is stronger than the evidence behind it. This pillar sets out what is actually grown, what each material type is measured to do, and why the thing a farmer receives is not a variety in the sense a buyer would understand the word.

Ghana grows the descendants of two separate cocoa introductions made about a century apart, and the first of them came in 1878, when Tetteh Quarshie brought West African Amelonado back from Fernando Po to the Akwapim hills¹, and West African Amelonado remained the planting material for the best part of a century. The second began with the Upper Amazon collections of the 1930s and 1940s, which the research station at Tafo crossed into the bi-parental hybrids COCOBOD distributes today.

A Ghanaian cocoa farmer planting a seedling in a shaded forest plot, other seedlings waiting in black nursery bags nearby, documentary photograph

The one decision you cannot revise

A tree crop's first decision sets the weight, the price and the yield of everything after it.

0.41-0.55 t/ha

Ghana's actual national yield, the reality check on every claim below

9 crosses

the official recommendation, every one carrying T85/799 as a parent

6 clones

the base population behind the breeding programme

What Ghana grows, and what it is called

The breeding programme behind the official recommendation works from a base population of six clones: P30, PA 7, PA 150, NA 79, T60/887 and T85/799³. One of those six, P30, is Amelonado itself, and in CRIG’s own analysis it returns the lowest contribution to yield efficiency of any parent analysed. Every claim below is read against Ghana’s actual national yield of 0.41 to 0.55 tonnes per hectare⁴. The boldest maturity claim, that current hybrids fruit in one and a half to two years⁵, is an agency statement reported by the Ghana News Agency, not a trial result.

Upper Amazon hybrid

Larger pods, a wider colour range, and the parent of today's official hybrids.

1.5-2 yr

the current maturity claim, an agency statement with no trial behind it

0.34-1.0 t/ha

the measured trial band a replant should actually be priced off

The maturity claim ladder: years to first pod for Amelonado, Upper Amazon progenies and current COCOBOD hybrids, and the evidence behind each — Figure 1 The maturity claim ladder: years to first pod, and the evidence behind each. Price a replant off the measured band of 0.34 to 1.0 tonnes, not off a brochure.

What this shows

The boldest claim rests on the weakest evidence. Amelonado’s 5-year maturity is a secondary citation of 2006 CRIG work; the Upper Amazon’s 3-year figure is a 1964 research-station potential, not a farm result; and the current hybrids’ 1.5 to 2-year claim is a 2022 agency statement with no trial behind it at all. Recommendations and measurements come from different studies in different years and are not paired observations.

Table 1: The planting materials, what each is claimed to do, and what kind of claim it is
Material	First pod	Yield claim	Disease	What kind of figure this is
West African Amelonado (P30, 'Tetteh Quarshie')	about 5 years	500 to 1,000 kg/ha	Lower losses than the hybrids over a 20-year trial	Secondary citation of 2006 CRIG work, via a 2022 thesis
Upper Amazon progenies	3rd year	about 2,000 kg/ha	Better CSSVD tolerance than Amelonado	A 1964 research-station potential. Not a farm result
Current COCOBOD hybrids (mixed hybrid seed)	1.5 to 2 years	Not stated as a figure	Symptom expression comparable to standards on inoculation	An agency claim by the Seed Production Division, 2022
Measured clone trial, Tafo (planted 2010)	n/a	0.34 to 0.54 t/ha; 0.7 to 1.0 best new clones	Comparable between new and standard clones	A measured field trial. The hardest numbers available
Ghana, national, actual	n/a	0.41 to 0.55 t/ha	n/a	The reality check on every row above

Sources: Asare (2022), citing Abdul-Karimu et al. (2006) and Toxopeus (1964); Padi et al. (CRIG, ICCO T1.30); Lockwood (1976); COCOBOD Seed Production Division via Ghana News Agency (2022); USDA Foreign Agricultural Service (2025). Recommendations and measurements come from different studies in different years and are not paired observations. Read the last column before using any figure in the third.

You do not buy a variety. You buy a population.

Five cocoa pods of different colours lined up on a woven mat, ranging from deep purple to yellow to orange, a farmer splitting a pod with a machete in the background, documentary photograph

One farm, five colours

No tree on a mixed-hybrid farm can be named to a cultivar.

West African Amelonado cocoa pods, small and deeply furrowed, yellow-green in colour, on a tree trunk with farm workers blurred in the background

West African Amelonado: small and deeply furrowed.

Upper Amazon hybrid cocoa pods, larger and a wider colour range from red to yellow, clustered on a branch

Upper Amazon hybrid: larger pods, a wider colour range.

The crosses themselves are controlled. Seed pods come from hand pollination of selected female clonal parents with pollen from specific male parents. What is not controlled is what a farmer ends up holding: CRIG’s own account is that most clones combine well with a few males for the development of mixed hybrids as planting material, so the pods issued are several crosses bulked together and not labelled by cross. Two farms planted from the same batch in the same week are not planted with the same thing, and neither can name what it has. That is what decides what a buyer can promise downstream.⁶

It also closes the premium most new entrants ask about first: in Ghana eight fine-flavour clones are in test and none has been released to Ghanaian farmers, and a single-origin or fine-flavour proposition cannot rest on material that is neither named nor traceable to a cross. That route is Pillar 7’s, built on documentation.⁷

The hybrid advantage is conditional, and the condition is the business

The case for hybrids is real but not free-standing. The account CRIG researchers report, from a 1998 consultancy assessment, is that hybrids bear in three years against at least five and carry more pods per tree, with a condition attached: they require optimal weather, chemicals, pruning, spraying and frequent year-round harvesting⁸. Planted without those, a hybrid is an expensive Amelonado. The longest side-by-side comparison run in Ghana cuts both ways: the Upper Amazon hybrids held a yield advantage over Amelonado across the full twenty years, yet disease losses were lower on Amelonado than on any of the seven hybrids tested alongside it⁹.

Adoption follows the condition attached to the hybrids rather than the yield claim made for them, and the survey evidence bears that out. A 300-farmer survey across all cocoa regions, published in 2012, put hybrid uptake at 44.0 per cent of farmers, with fertiliser at 33.0 per cent and black-pod fungicide at 7.5. The material is free and more than half of farmers were still not using it, which puts the constraint in the package around the seedling rather than in the seedling.¹⁰

44.0%

hybrid uptake on a 300-farmer survey, 2012

7.5%

black-pod fungicide use in the same survey

Opportunity

Sell the package, not the seedling. The state gives the planting material away, so a nursery competes on availability and provenance rather than on price. The margin is in what makes the material perform, and on this playbook’s own evidence that means fungicide timed to the pod cycle, which Pillar 4 finds one of only two practices that narrowed the yield gap, plus a written record of what was planted where. Farmers already state a willingness to pay for both material and rehabilitation, and adoption of a free input has stalled at 44 per cent, so the constraint being sold against is documented.

Where the material comes from, and the queue for it

A worker watering rows of cocoa seedlings in black nursery bags under a thatched shade structure at a seed garden, documentary photograph

A single concentrated source

90.4 million seedlings, 481 nursery sites, one season.

83.1%

of farmers source seedlings from COCOBOD seed gardens and cocoa stations

22.4%

of applicants could not get the variety they wanted

Planting material in Ghana comes from a single concentrated and public source. The Seed Production Division raised 90,367,452 seedlings at 481 nursery sites in 2020/21¹¹, and farmers draw on it overwhelmingly: 83.1 per cent from COCOBOD seed gardens and cocoa stations, 11.3 per cent from other farmers’ farms¹². The weakness is matching, not volume: 22.4 per cent could not get the variety they wanted. A free input a fifth of applicants cannot get in the form they want is a rationed input, and rationed inputs are allocated by proximity and relationship.

Two figures that an entrant usually asks for early both need care before they are used. There is no census of varietal composition: COCOBOD has published none, and the one share in circulation, about 13 per cent Amelonado, traces to a textbook with no survey behind it, so this playbook will not print it as a national figure.

And there is no official or transacted price for a seedling, because the state gives them away; what exists is what farmers say they would pay: GH¢0.36 a seedling, on a range of GH¢0.10 to GH¢2.00 (n=56), with 89.8 per cent of surveyed farmers said they would pay for the new varieties.¹³ The second weakness in the supply system is that it keeps no memory: the material is unnamed, so nothing on a Ghanaian cocoa farm records what was planted in which year, against an economic life of 25 to 30 years¹⁴, exactly the record a replanting decision needs.

The risks that sit inside the planting-material decision

A Ghanaian cocoa farmer holding a handful of mixed-colour cocoa pods beside a jute sack of dried beans, no visible provenance record, documentary photograph

Planting material of unknown origin

HIGH

What it is

Seed taken from a neighbour's farm or a roadside nursery with no seed-garden link. Second-generation hybrid seed does not breed true, so the vigour that justified the choice is not in the pod the farmer collects.

Evidence

About 11 per cent of farmers source planting material from other farmers’ farms, and Ghana’s recommended material is a set of named crosses that cannot be reproduced on-farm.

Who it hits

Any grower buying seedlings outside the Seed Production Division; any offtaker underwriting a replanting programme.

How to manage it, and the opening

Buy only against a seed-garden reference. Photograph and file the issue note. Where a block's origin cannot be evidenced, plan yield at the unimproved rate and price the risk.

A Ghanaian cocoa farmer inspecting a young hybrid seedling in a nursery bag, a ruled record book and ruler open on the ground beside him, documentary photograph

Paying hybrid prices for Amelonado performance

MEDIUM

What it is

A hybrid planted without the shade, pruning, fertiliser and spray regime it was bred for returns Amelonado yields on a hybrid cost base, and the shortfall gets blamed on the material rather than the package.

Evidence

The conditions come from a 1998 consultancy assessment (MASDAR). Hybrid adoption reached 44.0 per cent while fungicide use reached only 7.5 per cent.

Who it hits

Entrepreneurs modelling a replant; investors underwriting a yield uplift.

How to manage it, and the opening

Underwrite the input package and the labour calendar in the same cheque as the seedlings, or model yield at the unimproved rate. Pillar 4 sets out the regime and Pillar 5 what it costs.

Openings: where the planting-material gap pays

01

Sell the package, not the seedling. The state gives the planting material away, so a nursery competes on availability and provenance rather than on price. The margin is in what makes the material perform: fungicide timed to the pod cycle, plus a written record of what was planted where. Farmers already state a willingness to pay for both material and rehabilitation, and adoption of a free input has stalled at 44 per cent, so the constraint being sold against is documented.

02

Sell the provenance note, not a better tree. The supply system keeps no memory: nothing on a Ghanaian cocoa farm records what was planted in which year, against an economic life of 25 to 30 years, exactly the record a replanting decision needs. A nursery issuing a written provenance note carrying the seed garden, the cross where known, and the issue date is not selling a better tree; it is selling the only document that can date the block.

Key takeaways

01

Ghana grows a mixed hybrid population, not a named variety. Nine crosses are recommended and six clones dominate the breeding base, but what reaches a farm is several crosses bulked as mixed hybrid seed that nobody can name afterwards.

02

The maturity claims run 5 years, 3 years and 1.5 to 2 years, and they are a secondary citation, a 1964 station trial and an agency statement. Only the last is current and it is the least evidenced.

03

Measured trial yields are 0.34 to 1.0 t/ha against a national actual of 0.41 to 0.55. That distance belongs to Pillar 4, not to the material.

04

The hybrid advantage is conditional on an input and labour package. Adoption stalled at 44 per cent on a free input, so the binding constraint is the package.

05

No fine-flavour clone has been released in Ghana. A specialty premium has to be built on documentation, not genetics.

Written for each reader

Practitioner intelligence

Hover any card to pause and lift it.

For students

If you are thinking of entering cocoa, do not start by acquiring a farm. Start by learning to tell planting material apart in the field, because almost nobody can and every replanting decision depends on it. Spend a season with a seed-garden issue note in one hand and a farm in front of you, recording what was planted, when, and what it now yields. First-timers assume the seedling is the investment; it is free, and the money is in everything that follows it. You come out with a scarce skill and a dataset nobody else holds.

For entrepreneurs

Secure the offtake and the agronomy contract before buying a single seedling. The planting material is free from the Seed Production Division, so there is no business in supplying it; the business is the package that makes it perform, sold per hectare per season. Build the service around what the adoption data says is missing and what Pillar 4 finds actually moves yield: fungicide timing and a written record. The trap is quoting a yield uplift off a brochure; quote off the measured band, and only where you control the inputs.

For investors

The thesis is that Ghana's yield problem is not genetic, so any deal priced on a varietal uplift is mispriced. Three diligence asks: what is the documented provenance of the existing block and can the seller produce issue notes; what is the tree age profile against the 25 to 30 year economic life; and is the input package contracted or merely assumed. Release against a planting record and a first-harvest audit, not against hectares planted. The dominant risk is that unnamed material leaves the asset's age unverifiable at exit: take that discount at entry.

For ecosystem actors

The lever is a provenance record on every batch the Seed Production Division issues, carrying the seed garden, the cross where known, and the issue date. It costs little; the distribution channel exists. The measurable outcome is the share of the national estate whose planting year can be evidenced, now close to nil and needed by every rehabilitation programme to target. The failure to avoid is funding another seedling-raising target: volume is not the constraint, matching and memory are.

For students

If you are thinking of entering cocoa, do not start by acquiring a farm. Start by learning to tell planting material apart in the field, because almost nobody can and every replanting decision depends on it. Spend a season with a seed-garden issue note in one hand and a farm in front of you, recording what was planted, when, and what it now yields. First-timers assume the seedling is the investment; it is free, and the money is in everything that follows it. You come out with a scarce skill and a dataset nobody else holds.

For entrepreneurs

Secure the offtake and the agronomy contract before buying a single seedling. The planting material is free from the Seed Production Division, so there is no business in supplying it; the business is the package that makes it perform, sold per hectare per season. Build the service around what the adoption data says is missing and what Pillar 4 finds actually moves yield: fungicide timing and a written record. The trap is quoting a yield uplift off a brochure; quote off the measured band, and only where you control the inputs.

For investors

The thesis is that Ghana's yield problem is not genetic, so any deal priced on a varietal uplift is mispriced. Three diligence asks: what is the documented provenance of the existing block and can the seller produce issue notes; what is the tree age profile against the 25 to 30 year economic life; and is the input package contracted or merely assumed. Release against a planting record and a first-harvest audit, not against hectares planted. The dominant risk is that unnamed material leaves the asset's age unverifiable at exit: take that discount at entry.

For ecosystem actors

The lever is a provenance record on every batch the Seed Production Division issues, carrying the seed garden, the cross where known, and the issue date. It costs little; the distribution channel exists. The measurable outcome is the share of the national estate whose planting year can be evidenced, now close to nil and needed by every rehabilitation programme to target. The failure to avoid is funding another seedling-raising target: volume is not the constraint, matching and memory are.

Where this connects. Pillar 4 owns the planting regime and the yield gap; Pillar 3 the disease pressure variety tolerance is measured against; Pillar 7 the documentation route to a premium this pillar closes off on genetic grounds.

Footnotes

Emmanuel Asare, Morphological Characterization of Cacao Cultivars in Different Socio-Ecological Settings of Ghana (MSc thesis, Georg-August-Universität Göttingen and University of Kassel; hosted by the CocoaSoils programme, 2022), https://cocoasoils.org/wp-content/uploads/2024/01/Morphological-characterization-of-cacao-cultivars-.pdf.↑
Andrew Daymond et al., A Global Review of Cocoa Farming Systems (Reading, UK: University of Reading; funded by the International Cocoa Organization and the Swiss Foundation of the Cocoa and Chocolate Economy, 2021).↑
Francis K. Padi et al., Potential of Recurrent Selection for Developing Improved Cocoa Varieties in Ghana (Conference paper T1.30. London: International Cocoa Organization, n.d.), https://www.icco.org/wp-content/uploads/T1.30.-POTENTIAL-OF-RECURRENT-SELECTION-FOR-DEVELOPING-IMPROVED-COCOA-VARIETIES-IN-GHANA.pdf.↑
Joshua Taylor, Mariano J. Beillard, and Jeffrey D. Galloway, Ghana: Cocoa Sector Overview 2025 (USDA Foreign Agricultural Service, Global Agricultural Information Network, GAIN Report No. GH2025-0008, 2025), https://apps.fas.usda.gov/newgainapi/api/Report/DownloadReportByFileName?fileName=Ghana+-+Cocoa+Sector+Overview+-+2025_Accra_Ghana_GH2025-0008.pdf.↑
Ghana News Agency, COCOBOD Raises 140 Million Cocoa Seedlings to Increase Yields (Accra: Ghana News Agency, 2022), https://gna.org.gh/2022/05/cocobod-raises-140-millon-cocoa-seedlings-to-increase-yields/.↑
Taylor, Beillard, and Galloway, Ghana: Cocoa Sector Overview.↑
Daymond et al., Global Review of Cocoa Farming Systems.↑
Francis Aneani and Francis Padi, "An Assessment of Farmers' Adoption of Cocoa Rehabilitation Techniques in Ghana," Sustainable Agriculture Research 6, no. 1 (2017): 13-23, https://ageconsearch.umn.edu/record/265129/files/P2-p13-23.pdf.↑
G. Lockwood, "A Comparison of the Growth and Yield during a 20 Year Period of Amelonado and Upper Amazon Hybrid Cocoa in Ghana," Euphytica 25 (1976): 647-58, https://doi.org/10.1007/BF00041602.↑
F. Aneani et al., "Adoption of Some Cocoa Production Technologies by Cocoa Farmers in Ghana," Sustainable Agriculture Research 1, no. 1 (2012): 103-17.↑
Ghana Cocoa Board (COCOBOD), 52nd Annual Report and Consolidated Financial Statements for the Year Ended 30 September 2021 (Accra: Ghana Cocoa Board, 2021), https://cocobod.gh/resource_files/52nd-annual-report-and-financial-statements-2020-2021.pdf.↑
Aneani and Padi, "Farmers' Adoption of Cocoa Rehabilitation."↑
Aneani and Padi, "Farmers' Adoption of Cocoa Rehabilitation."↑
Padi et al., Potential of Recurrent Selection.↑