Ghana grows more cassava than any other crop, and the harvest keeps climbing. That rise has come from both more land and higher yield, yet yield still runs far below its ceiling: the national average is 24.27 tonnes a hectare against the 45 the same varieties reach on station, and one district already clears 43. The gap is not a mystery of climate or genetics. It is clean planting material and the variety a farmer plants, and each of those is a business someone can build; the field agronomy that backs them is Pillar 4.
Cassava is Ghana’s largest food crop by weight, and the crop most households name as their fallback when money or rain runs short. National output climbed from 17.2 million tonnes in 2015 to 28.5 million in 2024, a rise of about two thirds in a decade. That growth came from two things at once: farmers planted more land, from roughly 917,000 to 1,165,000 hectares, and they also lifted yield per hectare by about a third, from 18.8 to 24.27 tonnes. Both matter, but the second is where the durable money is, because land eventually runs out and yield does not.1
Even after that rise, the national average is only 54 percent of what the crop can give. The best district, Sekyere East in Ashanti, already averages 43.38 tonnes a hectare, which proves the ceiling is reachable on Ghanaian soil, not just on a research plot. Production also sits in a clear belt, with Eastern, Ashanti and Central regions leading and Bono East and Bono behind them, all in the forest and transition south. For anyone entering, the yield gap is the opportunity: closing part of it on a given farm needs no new land, no new variety science and no imported machine, only better planting material, timely weeding and soil that is fed rather than mined.2
Output rose faster than either input alone, so both more land and better farming are at work.
The national average sits at just 54 per cent of the ceiling the same varieties already reach. One district, Sekyere East, proves it is reachable on Ghanaian soil, not only on a research plot.
Ghana has a deep bench of released varieties. Research and development since 1993 has released dozens of them, and on average they yield about three times the local landraces they replace.3 The workhorse is Afisiafi: 28 to 35 tonnes a hectare, 32 percent dry matter, tolerant to cassava mosaic virus, and useful for starch, gari and flour. The high-yield names go further, with CRI-Agbelifia rated at a potential 50.8 tonnes and CRI-Bankye Hemaa at 48, the latter with the baking quality a flour buyer wants.4
Yield is not the only axis, and choosing on yield alone is the classic first-timer mistake. Tek-Bankye yields a strong 30 to 40 tonnes but is susceptible to cassava mosaic virus, so it is the wrong plant anywhere the disease is present, whereas Afisiafi and Nkabom trade a little yield for the mosaic tolerance that keeps a crop standing.5 The nutrition angle sits in the newer wave: yellow-fleshed, high-beta-carotene, mosaic-resistant lines released by CSIR-Crops Research Institute in 2022 and by the University of Cape Coast in December 2024, bred to put pro-vitamin A into a staple that usually carries none.6
| Variety | Yield (t/ha) | Dry matter | CMD status | Best market use | The trade-off to weigh |
|---|---|---|---|---|---|
| Afisiafi | 28 to 35 | 32% | Tolerant | Starch, gari, flour | The dependable all-rounder; not the top yielder, but forgiving to grow |
| CRI-Agbelifia | up to 50.8* | 24% starch | Not stated | Starch, gari | Highest yield on the list, but needs good management to reach it |
| CRI-Bankye Hemaa | up to 48* | 21% starch | Not stated | Fufu, bakery | High yield and bakery quality, but lower dry matter than starch buyers want |
| Tek-Bankye | 30 to 40 | 30% | Susceptible | Fufu, gari, ampesi | Strong yield, but a poor choice anywhere mosaic disease is present |
| Nkabom | 28 to 32 | 32% | Tolerant | Starch, fufu | A reliable, disease-tolerant all-rounder; mid-range yield with dependable quality |
Sources: * Potential (on-station) yield. Sources: CSIR-CRI (2017); MoFA/NVRRC (2019). CMD status shown only where the registry states it.
A starch or ethanol buyer pays for dry matter, so Afisiafi and the high dry-matter lines win. A flour or bakery buyer wants Bankye Hemaa. A nutrition or institutional-feeding buyer wants a yellow-fleshed line. Match the variety to the offtake before the first cutting goes into the ground. Deciding once the harvest is already on the ground looking for a buyer is too late.
Put the choice the other way round and it gets simpler. A variety is not chosen for its own qualities; it is chosen for a buyer, and the question is which cassava the person paying wants, and how long you have to get it to them. Table 2 reads the decision from the buyer down.
| If your buyer is | What they are buying | What that demands of the variety | Where it points |
|---|---|---|---|
| The household pot, through the market | Fufu, ampesi, boiled root | Poundability and taste; dry matter matters less than texture, and the root must be eaten within days | The default market, the lowest unit price, and the sale that punishes a slow seller hardest |
| A gari processor | Bulk roots, converted at roughly a quarter of their fresh weight | Volume and a steady supply more than any single quality trait; bitter types are welcome because processing is part of the product | The mass engine of the chain, and the tier with real Ghanaian profit data (Pillar 5) |
| A starch or flour plant | Dry matter, which is what they actually pay for | The highest dry matter available and a harvest window matched to the plant rather than the farm; fertiliser shape matters here most (Pillar 4) | The industrial cohort, bred for exactly this and adopted by almost nobody |
| An ethanol producer | Fermentable starch, at scale, close to the plant | Distance to the plant matters more than the variety | One buyer, one plant, and a large import bill behind it (Pillar 2) |
| Your own field, next season | Planting material | Disease-free stems of a known age; a different product from a root, and it should be priced as one | The business almost nobody is in, and the reason the yield gap persists |
Sources: CSIR released-varieties catalogue (2017); Acheampong et al. (2021); International Atomic Energy Agency (2018). Buyer-side economics are developed in Pillars 2 and 5.
The variety also decides how many days a farmer has to sell, which cassava treats as biology and a business should treat as logistics. A root begins to spoil within 24 to 72 hours of harvest, but how fast is set by the genotype, and deterioration-tolerant lines hold far longer than susceptible ones; Pillar 3 carries that clock in full. The catch is a genuine trade-off: post-harvest deterioration rises with root dry matter and falls with carotenoid content, so the high dry-matter roots a starch or flour plant pays most for are the very ones that rot fastest, while the longest-keeping roots tend to be worth less to the industrial buyer (Dankwa et al., 2025).
The published variety information covers yield, dry matter and disease resistance, but not how many days a root will hold, so a farmer choosing a stem cannot price the window. For an aggregator or a processor, matching a long-keeping variety to a long haul, and a high dry-matter, fast-spoiling variety to a plant within an hour’s drive, is free margin, available now and needing no new technology and no new variety.
The science is settled; the catch is reach. Only about 40 percent of farmers grow improved varieties at all; more than half the national crop is still lower-yielding local or farmer-saved material. That single fact explains much of the yield gap, and it points straight at the planting-material business.7
Cassava is grown from stem cuttings rather than seed. Healthy planting material means 15 to 20 centimetre cuttings taken from vigorous mother plants 8 to 12 months old.8 The trouble is that the same cutting which multiplies a variety also multiplies its diseases. Cassava mosaic disease travels inside infected stems, so a farmer who saves cuttings from a sick field plants the virus straight into the next one. Mosaic is already widespread in Ghana and cuts yields by more than 20 percent, depending on when infection strikes and which viral strains combine.9
Ghana has the institutions for clean seed. The Seed Inspection and Certification Division, under the Plant Protection and Regulatory Services Directorate, certifies foundation and certified seed and primary and secondary planting materials, cassava included.10 What is not yet on the public record is a dedicated clean-cassava-seed scheme at scale, or current figures on how much certified planting material actually reaches farmers. That gap is the opening.
A registered nursery that multiplies clean, disease-tested cuttings of a named variety and sells them to farmers on an offtake contract sits exactly where the yield gap and the disease risk meet. It is asset-light, it is repeatable season after season, and it is the single highest-leverage entry point in this pillar. Whoever builds the clean-seed layer captures value from every hectare planted downstream of it.
What a stem business actually sells is not the stem; it is the disease status and the age of the mother plant, and a cassava stick multiplies roughly eightfold in a year, so one clean field becomes eight fields’ worth of clean material in twelve months (International Atomic Energy Agency, 2018). That is why the missing seed market is a business that can pay: the biology is easy, but nobody is yet paid for the clean stem. Table 3 sets out three ways in.
| The entry | What you actually sell | What it costs to start | The trap |
|---|---|---|---|
| Clean stem multiplication | Disease status and a known mother-plant age rather than the stem itself | Land, clean starter material and a year; the material compounds roughly eightfold a year | A market with free dirty stems at every field edge; you need a buyer who loses money on disease, which usually means a processor rather than a farmer |
| Variety-to-buyer advisory | The match between a variety, a buyer's specification and a distance | Almost nothing; the information gap is the asset | Nobody pays for advice they can ask a neighbour for free; bundle it with the stem or the offtake, never sell it alone |
| Aggregating clean-planted farms | A predictable root supply with a known window and known dry matter | Working capital and a relationship rather than equipment | Contracting on price alone; the evidence says price does not hold growers (Pillar 6) |
Sources: Poku, Birner and Gupta (2018) on the missing seed market; Ministry of Food and Agriculture (2025) on certification; International Atomic Energy Agency (2018) on multiplication. Start-up costs are indicative and require field validation.
One threat is worth watching. Cassava brown streak disease, the more destructive cousin of mosaic, is not yet present in Ghana as of September 2025, but it is spreading westward across Africa and its main entry route is infected stems.11 Anyone moving planting material across borders is the risk. Anyone building a tested clean-seed system is the defence.
Carried inside planting material and already widespread, mosaic cuts yields by over 20 percent and travels with every saved cutting from an infected field.
Mosaic is already widespread in Ghana and cuts yields by more than 20 percent, depending on when infection strikes and which viral strains combine. Cassava brown streak disease, its more destructive cousin, is not yet present as of September 2025 but is spreading westward.
Every smallholder who replants from last season's own field, and every programme paying for a yield gain it will not get if the material carries disease.
Plant certified or tested cuttings of tolerant varieties such as Afisiafi; rogue out and destroy diseased plants; never take stems from a sick field.
Build the clean-cutting nursery. A registered nursery that multiplies clean, disease-tested cuttings of a named variety and sells them to farmers on an offtake contract sits exactly where the yield gap and the disease risk meet. It is asset-light, it is repeatable season after season, and it is the single highest-leverage entry point in this pillar. Whoever builds the clean-seed layer captures value from every hectare planted downstream of it.
Sell shelf life nobody else prices. The published variety information covers yield, dry matter and disease resistance, but not how many days a root will hold, so a farmer choosing a stem cannot price the window. Matching a long-keeping variety to a long haul, and a high dry-matter, fast-spoiling variety to a plant within an hour's drive, is free margin, available now and needing no new technology and no new variety.
Match the variety to the buyer, not the yield line. The match between a variety, a buyer's specification and a distance is an asset in itself, and almost nothing to start. Nobody pays for advice they can ask a neighbour for free, so bundle it with the stem or the offtake, and never sell it alone.
Output is rising, but at 24.27 t/ha the national average is only 54 percent of the 45 t/ha the same varieties reach on station. The gap is the opportunity.
Only about 40 percent of farmers grow improved varieties; the rest of the crop is lower-yielding local or saved material.
Choose variety by the market you are selling into: high dry-matter lines for starch, Bankye Hemaa for flour and bakery, yellow-fleshed lines for nutrition contracts.
Clean planting material is the highest-leverage entry point, because it closes the yield gap and blocks disease in the same move.
The variety also decides how many days you have to sell: some lines spoil within a day of harvest while others hold for a week or more, and that window is set at planting.
Where this connects. Variety and dry matter decide which demand channel a crop can serve and which product it becomes (Pillars 2 and 6). The yield gap and farm-gate economics set the cost base and the returns model (Pillar 5). And planting-material supply is the first weak link in the value chain and the clearest missing-middle opening (Pillar 6).