Ethiopian landraces: the coffee without a name

Pick up a bag of Colombian coffee and it'll usually tell you the variety: Castillo, Caturra, Pink Bourbon. Pick up a bag from Ethiopia and, more often than not, it says something vaguer. Heirloom. Landrace. Indigenous varieties. Ethiopian wild.

There's a good reason for the vagueness, and only occasionally a marketing one. Ethiopia is where arabica began and diversified. Much of the arabica grown elsewhere descends from a narrow slice of Ethiopian diversity that passed through Yemen. Ethiopia still holds a far broader spread of wild and cultivated coffee, much of it never individually catalogued and often growing side by side on the same hillside.

This piece covers what those labels actually mean, where Ethiopia's numbered JARC varieties fit in, and why an honest Ethiopian coffee sometimes can't name its variety at all. Short version: "landrace" describes how coffee lives in Ethiopia, and it covers several quite different things.

Heirloom, landrace, wild: what the names mean

The terms overlap, and the industry has been drifting from one to the next for good reason.

Heirloom was the default for years. It's borrowed from horticulture, where it loosely means an old variety passed down through generations. In other crops it can mean anything from "grown before the Second World War" to "handed down in a family", so it was always imprecise. In 2018 Square Mile announced it would stop using "heirloom" for its Ethiopian coffees, prompted by Getu Bekele and Timothy Hill's A Reference Guide to Ethiopian Coffee Varieties. The book argued that the word also implies nothing in Ethiopia has been studied or improved, which is plainly untrue.

Landrace is now the term most specialty roasters and coffee writers reach for. A landrace is a traditional, locally adapted population of a crop, shaped by its environment and by generations of farmers saving seed rather than by formal breeding. Bekele, speaking to Sprudge, put it neatly: "Landrace is not bounded by the concept of time and ownership." He also concedes it's still a catch-all.

Wild or forest coffee means coffee regenerating naturally in native forest, with little or no deliberate human selection. It's real, but it's a small slice of what's exported (more on production systems below). Indigenous or local is broader: it can include native Ethiopian material that's cultivated, selected or locally named.

Read the first four as broad descriptions of provenance rather than a single formal cultivar. The last one gives you specific information about a released selection.

Where arabica began, and why the rest of the world has so little of it

Arabica is itself a natural hybrid. A 2024 genome study in Nature Genetics found it formed when two other coffee species, Coffea canephora (robusta) and Coffea eugenioides, crossed and doubled their chromosomes. The researchers put that founding event at roughly 350,000 to 610,000 years ago, in the Ethiopian region, and date the split between wild populations and the ancestors of cultivated coffee to around 30,500 years ago.

Arabica started out with little diversity and lost more through several population crashes long before anyone farmed it. Ethiopia kept what was left: a much broader share of the species' genetic diversity, in wild forest populations and cultivated coffee, than anywhere the crop later travelled. Then people narrowed it further, twice.

First, a relatively small number of seeds crossed the Red Sea to be farmed on Yemen's terraces. A 2020 genetic study confirmed Yemen as arabica's secondary dispersal centre: almost all coffee grown outside Ethiopia descends from those early Yemeni farms.

Then a small founder population left Yemen for India, Java, Réunion and Europe's botanic gardens. Those lineages became the Typica and Bourbon families, the ancestors of many classic varieties grown across Latin America, Asia and beyond.

Work published in Agronomy in 2022 traced where those first seeds came from. The genetic group linking Ethiopia to Yemen came from the south, around Gedeo, Guji and Sidama, where Yirgacheffe and Guji coffees are grown today. Two other core Ethiopian clusters, largely from the west and south-west, never took part in domestication at all. The same study suggests Harar's coffee is partly domesticated coffee reintroduced from Yemen.

So when you drink a Caturra from Colombia, you're drinking from a narrow branch of the family. A southern Ethiopian coffee puts you much nearer the trunk.

How coffee actually grows in Ethiopia

Only a small share of Ethiopian coffee is truly wild, and most of the rest grows in ways you won't see in other coffee countries. Ethiopia is usually described as having four production systems, on a sliding scale from untouched forest to managed plantation. Published shares vary by source and year, so treat these as rough ranges.

Sources: Partnerships for Forests, Melbourne Coffee Merchants. Shares vary by source, year and region.

The middle two rows matter most for varieties. In semi-forest and garden coffee, the trees are a mix of whatever regenerated from the forest, whatever the family saved seed from, and whatever seedlings a nursery or government programme handed out. Nobody planted them in labelled rows. A typical garden plot is well under a hectare, and a washing station may buy cherry from hundreds or thousands of them.

Three layers inside an Ethiopian lot

It's natural to picture "landrace" as one shared pool of wild coffee. In practice, an Ethiopian lot can draw on three distinct layers of genetics, and most export lots draw on more than one.

Unless a producer separates trees or varieties at picking and processing, more than one layer can end up in the same cherry delivery. In that case, a broad label such as "landrace" is more honest than a single variety name.

Wild forest coffee is the uncatalogued layer: self-sown populations in the montane forests of the south-west, such as Kaffa, Sheka and Yayu, and in pockets elsewhere such as the Harenna forest in Bale. Neighbouring trees can be genetically quite different.

Regional landraces are what happens when people live with that coffee for centuries. Arabica mostly pollinates itself, so seed saved from a good tree tends to come back much like its parent. Over generations, farmers in each area have favoured particular types, and some of those have local names, often borrowed from native trees.

JARC selections are the formal-improvement layer, and the one most often misunderstood. They're Ethiopian through and through. Most are pure-line selections of individual Ethiopian trees, made by the Jimma Agricultural Research Center and released under catalogue numbers. A 74110 still comes from Ethiopian landrace germplasm. It simply has a name, a documented mother tree and a formal release history.

That clears up a common question. Only formally released selections and hybrids have JARC names or numbers, so the system can't be used to classify Ethiopian coffee in general. JARC reported about 44 released varieties by 2023, a small, formal subset of Ethiopia's much larger store of coffee genetics. Much of Ethiopia's coffee has no catalogue entry at all.

The JARC story: from crisis to catalogue numbers

Ethiopia's numbered varieties exist because of a fungus. Coffee berry disease (CBD), which rots green cherries on the branch, was first reported in Ethiopia in 1971 and became a major economic threat. On local landraces, national average losses have been estimated at 24 to 30% of the crop.

The response came from the Jimma Agricultural Research Center (JARC), founded in 1967 and now Ethiopia's national hub for coffee research. A full breeding programme began in the early 1970s, straight after the CBD outbreak. Researchers screened trees on forests and farms for disease resistance and other useful traits, propagated the most promising, and tested the selections for yield and cup quality.

Sources: Lemi Waya, JARC, ASIC 2023; landrace strategy paper, ASIC 2006.

Reading the numbers

The code is simpler than it looks, though the shorthand doesn't hold for every release. In the famous 74 series, the first two digits refer to the 1974 selection cohort and the last three identify the selection. It's the best-known and most widely distributed group of early CBD-resistant selections.

The selections you'll actually see

Sources: Tim Wendelboe, citing Bekele and Hill; Cup of Excellence Ethiopia 2021; Cup of Excellence Ethiopia 2024.

These varieties are genetically more varied than the numbering suggests. A 2021 study of 40 commercial JARC varieties found high genetic diversity and relatively low similarity between most pairs. And while the early selections were chosen mainly for disease resistance and yield, with cup quality assessed alongside, their competition results show how far the underlying Ethiopian material can go in the cup.

The landrace programme

From the 2000s, JARC and regional research centres turned their attention to local landrace development. Varieties selected in the humid south-west didn't always adapt well in Sidama, Yirgacheffe or Harar, and researchers wanted locally adapted material that would hold on to each region's performance and character. The 2006 strategy formalised the approach: collect local landraces in each region and select from them through a network of research centres at Jimma, Agaro, Gera, Tepi, Awada, Haru and Mechara.

The Awada centre, which serves Sidama and Gedeo, has released four such varieties: Angefa, Koti, Fayate and Odicha. A later review counted 12 regional specialty varieties across Harerghe, Wollega, Sidama, Yirgacheffe and Limu. You'll rarely see these names on a bag, but they're clear evidence that Ethiopian coffee is selected and bred locally as well as gathered.

Regional landraces: Kurume, Dega, Wolisho and friends

In the south, especially Gedeo (home of Yirgacheffe) and Guji, farmers have their own names for the coffee they grow. These are the closest thing Ethiopia has to traditional, farmer-defined varieties, and they're turning up on bags more and more. Most are named after native trees whose fruit or habit the coffee resembles.

Sources: Burman Coffee Traders, Archers Coffee, Melbourne Coffee Merchants via Sample Coffee, Coffee Review.

Treat these names with affection but some caution. Nobody has yet shown that "Kurume" is a single genetic variety rather than a family resemblance: small, compact, small-fruited trees that farmers group together. Some JARC selections, including 74110 and 74112, are also called Kurume locally because they can look like farmer-described Kurume types. Producers may also combine different local types in a lot on purpose, so a label can reflect how a producer groups the trees as much as a strict genetic identity.

So a bag that says "Kurume, Dega" is telling you something real about the trees and the place, just less precisely than "Caturra" would.

Gesha: the landrace that left home

The best argument for what's hiding in Ethiopia's unnamed coffee is one of the most valuable coffee varieties in the world. Gesha is Ethiopian forest coffee that happened to be collected, labelled and taken abroad.

Seeds were gathered from the Geisha coffee area of south-west Ethiopia in 1936 by an expedition linked to the British consulate, according to consular documentation from that year. The material went to research stations in East Africa, where the spelling "Geisha" entered germplasm records. From there the Panamanian line reached CATIE in Costa Rica in 1953 and Panama in the 1960s, valued at first for its partial resistance to leaf rust rather than its flavour. In 2004 the Peterson family's Hacienda La Esmeralda entered it in the Best of Panama competition, where it won and sold for a record price at auction.

World Coffee Research notes that several genetically distinct plant types have been called Gesha, many from the same part of Ethiopia. The Panamanian line descends from one CATIE accession, T2722, and is comparatively uniform. Gesha is now planted back in its home region too, at farms such as Gesha Village, under its original name.

So the famous Panamanian Geisha is one named line of Ethiopian forest coffee, and an Ethiopian coffee labelled Gesha or Geisha may be a different type altogether.

Why we often can't name the variety

When an Ethiopian coffee says "landrace", it's usually for one or more of four reasons.

The farms are tiny and mixed. The great majority of Ethiopian coffee comes from smallholders. A garden plot may hold forest regenerants, trees grown from the family's own seed and a row of JARC seedlings from a nursery, all picked together. Nobody has a planting map.

The cherry is pooled. A washing station typically buys from hundreds of these farms, sometimes thousands, and processes the cherry together. Even if every farmer knew exactly what they grew, a washing station lot is a blend by design.

Traceability was reduced for a period. From 2008, most traded coffee moved through the Ethiopian Commodity Exchange, which graded and marketed coffee by region and quality and made farm-level traceability harder. There were exceptions, including cooperatives and some larger producers. Reforms from 2017 reopened direct export and traceability for eligible washing stations and exporters. Single-farm, single-variety lots became more feasible as a result, though they're still unusual.

The science is unfinished. There's no complete genetic catalogue of Ethiopian coffee, and most of it can't be reliably identified by eye. You'll see figures anywhere from 6,000 to 15,000 varieties quoted, but nobody has published a definitive national count. Local names may describe a look rather than a lineage. Even competition results carry oddities: the lot that placed third at the 2021 Cup of Excellence, from Guji, was listed simply as "Typica".

The most precise true statement about these coffees is usually about place and production: which zone, which kebele, which producer or washing station, what altitude, and how the lot was processed.

Does it matter in the cup?

Yes, though less than you might expect, and differently from elsewhere. In Latin America, knowing the variety tells you a lot because the range is narrow and well studied. In Ethiopia, the genetics are so varied and the growing conditions so particular that place and process often explain more of the cup than any variety name could.

Southern genetics and southern places go together. Gedeo, Guji and Sidama overlap strongly with the "Ethiopian Legacy" genetic group identified in the 2022 study. Those same places, their altitude and the way the coffee's processed all shape the florals and citrus buyers associate with classic Yirgacheffe, so it would be a stretch to credit that profile to genetics alone.

Regional character is worth protecting, and JARC's local landrace strategy was designed to do exactly that, favouring locally adapted material over one-size-fits-all varieties.

Named selections reach the very top. In the 2021 Cup of Excellence, eight of the top ten ranked lots were listed with JARC numbers, led by 74165, 74110, 74112 and 74158. In 2024, the first five ranked lots were all listed as 74158.

Mixed lots can be a deliberate choice. Some producers combine local types because the blend gives a more balanced, complete cup, so a variety list with two or three names can signal intent rather than poor separation.

A named variety is useful information, and it may point to a carefully separated lot. Cup quality still comes down to the producer, the place and the processing, whatever the label says.

Climate and conservation

Ethiopia's unnamed coffee is also one of arabica's most important genetic reserves. When coffee-growing countries need resistance to disease, heat or drought, Ethiopian material is where breeders look, because many cultivated varieties elsewhere descend from a narrow slice of the species' original diversity.

That reserve is under pressure. A 2017 study in Nature Plants, led by Kew, modelled Ethiopia's coffee lands to 2099. It found 39 to 59% of the current growing area could become unsuitable for coffee without major intervention, with parts of eastern Sidama among the more vulnerable areas. The same study found that moving coffee upslope into higher forest, combined with conservation, could substantially increase the area that stays climatically suitable.

The wild forests face more immediate threats too. Deforestation, land-use change and climate change, along with planting new varieties inside coffee forests, are driving genetic erosion and habitat loss.

There are protections. UNESCO designated Kafa and Yayu as biosphere reserves in 2010, and Sheka followed in 2012. Kafa alone covers about 760,000 hectares, much of it cloud forest, and holds one of Ethiopia's last large areas of montane forest. JARC and the Ethiopian Biodiversity Institute also keep large living collections of Ethiopian coffee.

The science is moving quickly. The 2024 arabica genome gives breeders a reference for studying Ethiopian diversity, and Ethiopian researchers are using DNA markers to fingerprint gene-bank collections and released varieties. Better identification should make it easier to tell named selections apart and to document local diversity. It also strengthens the case for paying prices that keep forest and garden coffee worth growing.

How to read an Ethiopian coffee label

A quick guide for the next bag you pick up.

And when a coffee says only "landrace", that's often the most honest label it can carry. It comes from the one country where arabica still grows as wild forest trees, farmer-selected landraces and numbered research selections, sometimes all on the same hillside.

Further reading

If you want to dig further, these are the five sources we'd start with.

• What Is Heirloom Coffee?, Sprudge (2022). Liz Clayton talks to Getu Bekele, co-author of A Reference Guide to Ethiopian Coffee Varieties, about why "landrace" fits Ethiopian coffee better than "heirloom", and why even landrace is still a catch-all.
• The genome and population genomics of allopolyploid Coffea arabica reveal the diversification history of modern coffee cultivars, Salojärvi et al., Nature Genetics (2024). The study behind our account of how arabica formed as a natural cross between robusta and Coffea eugenioides. Free to read, though heavy going in places.
• Deciphering Early Movements and Domestication of Coffea arabica through a Comprehensive Genetic Diversity Study Covering Ethiopia and Yemen, Montagnon et al., Agronomy (2022). Traces the coffee that left Ethiopia for Yemen back to Gedeo, Guji and Sidama, and suggests Harar's coffee is partly a return trip from Yemen. Free to read.
• Geisha (Panama), World Coffee Research. The variety profile behind our Gesha section, covering the CATIE accession T2722 and the several distinct plant types that have shared the Gesha name. Short and free.
• Resilience potential of the Ethiopian coffee sector under climate change, Moat et al., Nature Plants (2017). Kew's modelling of Ethiopia's coffee lands to 2099, and the source of the 39 to 59% figure. The abstract is free; the full paper sits behind a paywall.