414 guides across 21 subjects
Every figure traced to a named source
Prices in Ugandan shillings, per acre and per animal

Best Mango Varieties in Uganda

A ripening mango hanging from the tree

The highest yielding mango genotype ever measured in Uganda is also one of the most heavily attacked by fruit flies. The two cultivars the flies leave alone yield less. That single sentence is why lists of the best mango varieties for Uganda disagree with each other, and why choosing one is a trade rather than a lookup.

Uganda has one published mango yield trial, one published fruit fly susceptibility study, and no released mango variety at all. Read together, those three facts support a real ranking and refuse a simple one. What follows is the measured evidence, what it rules out, and how to pick against your own site and buyer.

Why No Single Mango Variety Is Best for Ugandan Farms

Three things make a mango variety worth planting, and no cultivar in the Ugandan record has all three.

The first is yield, and on the Ugandan evidence that is led by local material. The second is how little fruit fly damage it carries, and there the introduced cultivars win clearly. The third is fruit quality in the sense buyers mean: low fibre, a high ratio of flesh to seed, fruit that travels without spoiling. That is also where the introductions win, and it is why they were brought in.

So the local genotypes give you tonnage you can partly lose to flies and partly cannot sell to a processor. The introductions give you clean, saleable fruit at a lower yield per tree. Neither is a mistake. Which one is right depends on whether you are selling into a nearby fresh market that eats fibrous fruit happily, or trying to supply somebody with a sieve and a specification.

Measured Mango Yields by Variety at Namulonge

The trial behind these numbers evaluated 36 mango genotypes on a seven year old field at the National Crops Resources Research Institute at Namulonge, in randomised blocks with three replications and four trees per genotype in each, over three years. Trees were pruned twice a year and slashed monthly, one fruit fly trap was run per 30 metre radius, and no disease control was applied at all, on purpose. That last choice is what makes the figures usable for a variety decision: each genotype was left to hold its own crop without a fungicide programme propping it up.

Variety Kg per tree a year Recorded as Fruit fly load
Kate 59.6 Ugandan Among the highest
Suu 51.2 Ugandan Not tested
Glenn 49.6 Introduction High intensity
Tommy Atkins 48.7 Introduction Lowest
Vandyke 48.5 Introduction Not tested
Duncan 46.5 Introduction Not tested
Bire 44.2 Ugandan Highest incidence
Dodo red 39.5 Ugandan Middling
Kent 36.5 Introduction Low
Takataka 35.0 Ugandan Not tested
Apple mango 34.4 Ugandan Middling
Zillate 32.4 Introduction Low
Koona 31.6 Ugandan Not tested
Keitt 11.1 Introduction Highest intensity
Boribo 9.6 Ugandan High intensity
Alphonso 7.2 Introduction Not tested
Ngowe 4.0 Ugandan Not tested
Kensington pride 3.5 Introduction Not tested

Eighteen of the 36 are shown, chosen to cover the top, the middle and the bottom. The rest sit between: Kawanda wide 36.1, Heidi 35.7, Parvin 34.9, Florigon 34.1, Kawanda Green 32.8, R2E2 31.0, Julie 23.9, Sejjembe 18.3, Edward 14.0, Irwin 12.5, Asante 12.4, Pascal 11.9, MP1 7.5, Palmer 7.4, Pinero 5.5 and Haden 4.2 kilogrammes a tree. The average across all 36 works out at about 26 kg a tree a year.

Two things about how to read that column. It is a ranking of measured means, not proof that one variety beats the next: the variability statistics printed under the trial's yield table do not reconcile with each other, so the safe reading is the shape of the spread rather than a claim that, say, Suu reliably beats Vandyke. And these are annual means across three years, which matters because several of these genotypes bear in alternate years. Seven of the 36 produced nothing at all in one of the three years, among them Bire, which is otherwise a strong yielder.

The honest headline is the range itself. The best genotype yielded seventeen times the worst on the same field, in the same soil, under the same management. No fertilizer decision, no spray programme and no pruning regime available to a Ugandan grower moves yield by a factor of seventeen. Variety choice is the largest single lever in the orchard, and it is the one you cannot change afterwards.

The Fruit Fly Trade Off Behind Every Variety Recommendation

Ugandan advice on mango varieties is close to unanimous that the main commercial cultivars resist fruit flies. The Ugandan data says something different, and it is worth being precise about what was measured.

A study across three Ugandan agro ecological zones sampled twelve mango cultivars and incubated the fruit to count what came out. Overall, 35.0 percent of sampled fruits were infested, and 76.3 percent of those infestations were the same invasive Bactrocera fly, with the native species almost absent. Two measures were reported for each cultivar: incidence, the share of fruits with any infestation, and intensity, the number of larvae per kilogramme of fruit.

What the Ugandan fruit fly data actually shows by cultivar. Note that the two measures rank varieties differently, and that the rankings move between zones.
Lowest all round. Tommy Atkins and Kent. Tommy Atkins had the lowest infestation incidence in the Lake Victoria Crescent at 13.5 percent and Kent 19.5 percent, and both were also among the least infested by larval count, in the Crescent and in the north. These are the only two cultivars low on both measures in both zones.
Highest incidence. Biire at 44.7 percent, with Palvin next at 44.4, in the Lake Victoria Crescent. More than three times the Tommy Atkins figure.
Highest larval counts in the Crescent. Glen at 92.4 larvae per kilogramme and Boribo at 86.7.
Highest larval counts in the north. Keitt at 185.8 larvae per kilogramme and Kate at 125. Keitt was among the least infested cultivars in the Crescent on the same measure, which is how much zone matters.
Pooled across zones, the study puts Tommy Atkins and Kent as least infested and Keitt, Kate and Biire as most.
None of this is resistance. A fruit fly is an insect rather than a disease, and no mango cultivar in Uganda has been shown to resist it. What was measured is relative infestation, and even the best cultivar in the study carried flies in more than one fruit in eight.

Now lay the two studies side by side, which nobody appears to have done. Kate tops the yield table and carries one of the heaviest larval loads. Bire is seventh on yield and had the highest infestation incidence of the twelve cultivars tested. Tommy Atkins is fourth on yield and the cleanest on flies. Kent is ninth on yield and second cleanest. The trade is not subtle and it is not in the direction the popular lists suggest.

A widely repeated Ugandan list names Tommy Atkins, Apple mango, Kate and Alphonso as the four best varieties to plant and calls all four resistant to the fruit fly. On the Ugandan measurements, one of those four is genuinely low on flies, one is among the worst, one was mid ranked on flies and mid ranked on yield, and Alphonso yielded 7.2 kilogrammes a tree, fifth from the bottom of 36, and is an alternate bearer besides. The list is not a ranking of Ugandan performance.

Local Mango Varieties Against the Introductions

Here is the result that surprises people, including the people who wrote the promotional lists. Split the trial's 36 genotypes by how the trial itself records them and the fifteen Ugandan ones average 29.6 kilogrammes a tree while the twenty one introductions average 24.3. The local material out yielded the imports by about a fifth.

It is not a clean sweep. The top of the table mixes both: Kate and Suu are local, Glenn, Tommy Atkins, Vandyke and Duncan are introductions. And the very bottom is mostly introductions too, with Kensington pride, Haden, Pinero and Alphonso all under 8 kilogrammes a tree. What the averages rule out is the premise that improved imported cultivars are the high yielding option in Uganda. In the only Ugandan trial available, they are not.

Local types carry a second advantage that matters more on a low input farm than any yield figure. The local cultivars of East Africa include polyembryonic types, which the reference literature describes as less susceptible to anthracnose as a group, and anthracnose control in a wet season can call for up to twenty fungicide treatments. On a farm where nine growers in ten apply no fertilizer at all, a variety that needs no spray programme to hold a crop is worth more than a variety that yields more when sprayed.

There is a third difference, and it is a variety question rather than a nursery one. Some of the local cultivars are polyembryonic and breed true from seed, so a Dodo stays a Dodo and a Boribo stays a Boribo. Kent, Keitt, Tommy Atkins and Alphonso do not. A seedling of any of those four is an unknown tree wearing a known name, which means the identity of an introduced variety rests entirely on a graft that somebody else made and that you cannot inspect after the fact. The propagation side of this is covered on the mango growing guide; what matters for variety choice is that the two groups carry completely different levels of identity risk. A Makerere grafting trial at Kabanyolo makes the point neatly, putting Alphonso, Kent, Bire and Tommy Atkins scions onto a Kagogwa rootstock, which is a Ugandan landrace.

Fibre and Flesh to Seed: What Buyers Pay For

So why plant the introductions at all, if the local types yield more and need less? Because of what happens when the fruit reaches a buyer.

The Ugandan extension description of local mango is blunt. The flesh is fibrous, which consumers dislike because the fibre stays in the teeth, and which processors dislike because it blocks their sieves. And the flesh to seed ratio is poor: in local mango the quantity of flesh is roughly the size of the seed, while in varieties of high commercial value the flesh is five or six times the seed. The Namulonge researchers make the same point independently, noting that high fibre in most local cultivars makes them less attractive for processing and for export.

Follow that through and the local yield advantage shrinks. A tree carrying 44 kilogrammes of fruit whose stones account for a large share of the weight, and whose pulp a juicer will reject, is not delivering 44 kilogrammes of saleable product. Nobody has published a Ugandan flesh recovery figure by variety, so this cannot be turned into a number, but the direction is clear and it is the reason the varieties were introduced in the first place.

Grower preference reflects it. In a Makerere survey in western Uganda, Tommy was the most preferred variety on yield, taste, market, early fruiting and shelf life together, and Tommy, Kent and Bire were the three most grown. Preference is not measurement, but it is a market signal from people selling fruit every season.

Mango Variety Names Are Not Stable in Uganda

Before acting on any variety list, including the table above, it is worth knowing how unreliable mango names are in this country. This is not pedantry. It is the reason two published lists can recommend opposite things while appearing to agree.

Four naming problems that show up in Ugandan mango advice, each verified against the Ugandan research record.
Kate and Takataka are treated as one name and are two genotypes. A popular list writes "Kate (TakaTaka)" as though they were the same. The Namulonge trial lists them separately, at 59.6 and 35.0 kilogrammes a tree. Buying one believing it is the other is a 40 percent yield difference.
Kate and Keitt are not the same either. They are separate rows in the same trial, at 59.6 and 11.1 kilogrammes a tree, and separate entries in the fruit fly study, where Keitt carried the heaviest larval load recorded in the north. The names look alike in speech and the trees do not behave alike.
Spellings drift and so do the figures attached to them. Bire appears as Biire, Kagogwa as Kagoogwa, Dodo as Doodo, Sejjembe as Ssejjembe, Suu as Ssu. Circulating lists also carry Dancan for Duncan, Palmar for Palmer and Erwin for Irwin, which is how a figure gets attached to the wrong tree.
The origin column in the research does not mean what it looks like. Ugandan papers tabulate cultivar origins in ways that place well known Florida and Australian selections elsewhere. The explanation is in the trial's own prose: the collection consists of introductions from Kenya, South Africa and Puerto Rico plus Ugandan landraces, so the column records where Uganda obtained the material, not where the cultivar was bred. That is why this page says "recorded as an introduction" and prints no country of origin for any cultivar.
And the ground truth: in the western Uganda survey, 61.7 percent of mango growers could identify their varieties only by common name. When three growers in five cannot name the tree, a variety recommendation has nowhere reliable to land.

No Released Mango Variety and No Certified Seedling Supply

This is the part that changes how you should read every mango variety page, this one included.

Uganda's national research organisation publishes a list of varieties licensed to named companies for certified seed production, each carrying an identification number that appears on the packaging. That list runs to 84 entries across 12 crops: maize, beans, groundnuts, rice, finger millet, cow pea, sorghum, sesame, greengram, pigeon pea, sunflower and pearl millet. Every entry is an annual field crop. There is no mango on it, no fruit of any kind, and no vegetable.

So for mango there is no released variety, no licensed multiplier, and no certification mark to ask a nursery for. The mechanism exists and works, which is exactly how we know it is absent here rather than merely unpublished: a grower buying maize or bean seed can check a variety against that list and cannot do the same for a mango tree.

What does exist is a research collection. A Ugandan initiative gathered local material and brought in introductions from Kenya, South Africa and Puerto Rico, building a core collection of 38 accessions at Kawanda Agricultural Research Station which was later moved to Namulonge, held to supply germplasm for future improvement. A second Ugandan study drew its genotype information from the horticulture department at Namulonge and its fruit from a field collection at the Ngetta zonal institute in Lira. Those are conservation and evaluation collections. Evaluated is not released, and released is not multiplied.

The practical consequence sits in the same western survey. Only 16.7 percent of growers chose the varieties they planted. The remaining 83.3 percent waited to receive them, 63.9 percent from government and 22.7 percent from non governmental organisations, and the authors link the naming confusion directly to material being distributed without varietal names attached. Five growers in six planted whatever arrived. A page that tells you to plant a particular variety is describing a choice most Ugandan mango growers have never been offered.

Where the Official Best Variety List Breaks Down

Uganda's extension guidance carries its own variety recommendation, and it is a better starting point than the promotional lists because it is at least specific. Among the introductions it names Tommy Atkins, Zillate, Keitt, Kent, Parvin, Edward, Florigon and Glenn as the best performers under Ugandan conditions. Set that against the Namulonge trial and it holds up in part.

Glenn at 49.6, Tommy Atkins at 48.7 and Kent at 36.5 kilogrammes a tree earn their place. Parvin at 34.9, Florigon at 34.1 and Zillate at 32.4 are respectable and middling. But Keitt at 11.1 and Edward at 14.0 sit in the bottom third of 36 genotypes, and the same trial recorded both of them, along with Heidi, losing more than 45 percent of their set fruit to drop, which is what the authors identify as the cause of their low totals. Keitt also carried the heaviest fruit fly load measured anywhere in the Ugandan study. Two of the eight named best performers are contradicted by the country's own yield trial.

The same page's list of local selections is worth a look too, because it is cut off mid sentence in the published source, ending after three names and an "and". Uganda's official local mango variety list is incomplete where it is published, which is one more reason to treat the measured table above as the more useful document.

Choosing Mango Varieties for Your Site, Market and Input Level

Working rules that follow from the evidence rather than from a favourite.

Plant more than one, and the Ugandan data supports it from two directions. Growers who planted a mixture had higher production than those growing local types alone, and the highest count on a single farm in the western survey was nine varieties. A mixture spreads the harvest across weeks, spreads fruit fly pressure across cultivars with different loads, and covers you when an alternate bearer takes a year off. Since seven of the 36 genotypes produced nothing in one of three years, a single variety orchard is a single point of failure.

If your fruit goes to a processor or into any chain with a specification, lean on the low fibre introductions and accept the lower yield per tree. Tommy Atkins and Kent are the defensible pair: both are in the upper half of the yield table, both are the cleanest of the twelve cultivars tested for flies, and Tommy is the most preferred variety among Ugandan growers who sell.

If your fruit goes to a nearby fresh market that buys fibrous fruit without complaint, the local genotypes are a stronger proposition than their reputation. Kate, Suu and Bire led or nearly led the Ugandan yield table, they came true from seed in the polyembryonic cases, and they need less fungicide. Budget for fruit fly management instead, because Kate and Bire are among the worst for it.

Match the variety to your spray capacity rather than to a magazine. A cultivar that drops more than 45 percent of its set fruit, or that carries 185 larvae a kilogramme in your zone, is asking for an input programme. If you are among the nine growers in ten who apply nothing, plant the varieties that perform without help.

And whatever you decide, treat the label on the tree as a claim. Ask which rootstock was used, ask to see the mother trees carrying fruit, inspect the graft union before paying, and write down what you were told so that in five years you can tell whether you got it.

For the agronomy behind these choices, the guide to mango farming in Uganda covers site, altitude, nursery work and the establishment years, and mango orchard management covers the operations once trees are bearing. The wider fruit farming guides set mango beside avocado and passion fruit varieties, where the variety question has a different shape. Since no certified mango nursery scheme exists, the pages on choosing quality planting material and starting a seedling nursery are the closest thing to a defence, and dried fruit production is where the fibre and flesh to seed question turns into money.

Best Mango Varieties in Uganda: Common Questions

Which mango variety yields most in Uganda? Kate, at 59.6 kilogrammes a tree a year in the only Ugandan trial, against a trial average of about 26 and a bottom figure of 3.5. Two cautions before you plant it. Kate is also among the most fruit fly infested cultivars in the Ugandan data, at 125 larvae a kilogramme in the north. And the trial's own variability statistics do not reconcile, so treat the table as a ranking of measured means rather than proof that Kate beats Suu.

Which variety resists fruit flies? None. No mango cultivar in Uganda has been shown to resist fruit flies, and any page saying otherwise is overstating what was measured, which is relative infestation. Tommy Atkins and Kent are the two cultivars consistently at the low end across both measures and both zones tested, and even Tommy Atkins carried flies in more than one fruit in eight.

Are local mango varieties worse than imported ones? Not on yield. The fifteen genotypes recorded as Ugandan averaged 29.6 kilogrammes a tree against 24.3 for the twenty one introductions. Local types are worse on fibre and on flesh to seed ratio, which is a market and processing problem rather than a farming one, and better on anthracnose susceptibility, which is a farming one. Which matters more depends on who buys your fruit.

What does a grafted mango seedling cost? It varies by nursery, district and season, and there is no published Ugandan figure worth repeating. What is worth knowing is what you are paying for: one to two years of nursery work, a rootstock raised from local seed, a graft made at five or six months, and a scion whose identity you are trusting somebody about. A Ugandan nursery study found graft survival varying by cultivar, with the best of three at 70 percent, so a nursery that has thrown away its failures has costs to recover. Compare more than one supplier and price per surviving tree rather than per tree offered.

Is there a certified mango variety in Uganda? No. The national variety licensing list covers 84 entries across twelve annual field crops and contains no fruit at all. What exists for mango is a research germplasm collection that was assembled, moved and evaluated, not released. So there is no certification mark to ask for and no official variety register to check a nursery's claim against.

Can I grow mango from seed instead of buying grafted trees? For the local polyembryonic types, yes, because their seed comes true to the parent. Dodo and Boribo are the East African examples named in the literature. For the introductions it does not work: a Kent or a Tommy Atkins seed is a cross of unknown parentage, and you find out what you planted after several years of care.

How many varieties should one orchard carry? More than one, for three separate reasons: alternate bearing, since seven of 36 genotypes produced nothing in one of three years; fruit fly pressure, which differs sharply between cultivars; and harvest spread, since a single variety concentrates your crop into the weeks when everyone else's arrives and prices fall. Ugandan growers with mixtures reported higher production than those with local types alone.

Why do published Ugandan variety lists disagree so much? Partly because names drift, with Kate and Takataka run together and Bire spelled several ways. Partly because most lists are not built on measurement: the official best performer list names two cultivars that the country's own yield trial puts in the bottom third. And partly because most growers never chose a variety at all, so there is little accumulated farm experience to correct a bad list.

Does variety affect when the tree starts fruiting? Probably, and it is not measured in Uganda. The yield trial began on a seven year old field that was already bearing, so it says nothing about earliness, and "early fruiting" appears in the Ugandan record only as a grower preference ranking rather than a measurement. Treat any early bearing claim for a particular variety as unverified.

Prices for grafted trees and for fruit move with the season, the district and the variety, so ask locally and compare several nurseries before committing. The most valuable thing to ask for is not a price: it is to be shown the mother trees your scions were cut from, carrying fruit, before you pay for anything.

More in Fruit Farming