Liberica and Excelsa: India's Other Two Coffee Species
Liberica and Excelsa have stood on Indian estates since the late nineteenth century, mostly as boundary trees whose fruit nobody picked. One of them is a parent line of S795, India's most planted arabica. This guide covers what the two species are, why they stayed marginal, and what the 21 lots now in ICB's catalog actually are.
Most writing about Indian coffee stops at two species. Arabica grows on the higher estates, Robusta on the lower and wetter ones, and the differences between them, in the field and in the cup, are covered in ICB's guide to how Arabica and Robusta differ in India.
Two more species have been on Indian estates for about 150 years. Liberica arrived in 1872, and Excelsa came in the same wave of experiments. Neither became a commercial crop, and both were largely forgotten outside the estates where they still stand. As of 26 September 2026, ICB's catalog lists 15 Excelsa coffees and 6 Liberica coffees among 1,499 active listings, about 1.4% of the catalog combined.
Below: what the two species are, how they reached India, how one of them ended up inside most Indian arabica, why Excelsa is being planted again while Liberica is not, and what the current Liberica and Excelsa lots on the Indian market look like.
What is Liberica coffee, and what is Excelsa?
Liberica (Coffea liberica) is a coffee species from the lowland forests of West Africa, grown as a tall tree with large cherries and beans, resistant to leaf rust, and accounting for roughly 1% of world coffee. Excelsa (Coffea dewevrei) is a separate Central African species, farmed at mid elevations, with smaller beans and a dried-fruit cup. India grows both in small quantities.
A few years ago the second sentence of that definition would have read differently. From 2006, Excelsa was classified as a variety of Liberica (Coffea liberica var. dewevrei), and much of what is published online still says so. At least one Indian roaster listing in ICB's catalog describes its Excelsa as "classified under the Liberica variant."
Genomic work led by the Royal Botanic Gardens, Kew has since reversed that. Davis et al. (2025), published in Nature Plants, used genomic data to redraw the species boundaries within the Liberica group and confirmed Excelsa as a species in its own right. A follow-up paper (Davis et al., 2026, Scientific Reports) examines hybridisation between the two, which goes some way to explaining the long confusion: the two are close relatives and can cross.
Wild populations of Liberica have contracted sharply. Kew's 2019 assessment of wild coffee species (Davis et al., *Science Advances*) places many of the genus's species at risk of extinction, and breeders now pay closer attention to the cultivated trees that remain.
Info: The two species side by side. Liberica: native to upper West Africa (Sierra Leone to Nigeria); typical cultivated elevation 10–100 m; very tall tree; large beans; resistant to leaf rust. Excelsa: native to Central Africa (Congo basin to Uganda); typical cultivated elevation 500–1,500 m; tall tree; beans closer to arabica in size; late ripening, high yield, tolerant of leaf rust. Figures from the species profiles published by the South India Coffee Company (SICC), which also supplies Excelsa into the Indian market.
How they reached India, and why they stayed marginal
Both arrived because of leaf rust. The fungus was first reported on Ceylon's coffee estates in 1869 and went on to wipe out the island's coffee industry within two decades. Planters in South India, watching the same disease reach their own estates, went looking for coffee that could survive it.
Liberica was introduced to India in 1872. Coffee Board records associate its promotion with Colonel J. R. Benson, a planter who argued for it as a replacement for arabica on rust-affected estates. It drew attention through the 1870s and 1880s for its apparent rust tolerance and vigorous growth; Edmund Hull's 1877 planting manual for southern India and Ceylon records the interest. Excelsa appeared in Indian estate records over the following decades, in the same search for rust-tolerant material.
Neither held up as a crop. Liberica trees regularly grew past ten metres, which made pruning and harvesting expensive. Cherries ripened unevenly on the same branch, and the cup was inconsistent. When planters needed a hardier alternative to arabica, they turned to Robusta, which was shorter, easier to manage and more predictable. By the early twentieth century most planters had made that switch.
The trees were not removed. Liberica and Excelsa stayed on as boundary trees, windbreaks, shade and grafting rootstock, where deep roots and disease tolerance had practical value. Mercara Gold Estate in Coorg still calls Liberica "tree coffee," a name it says is used on most plantations. The German importer Coffee Consulate reports that some Indian regions refer to Liberica as a "sacrifice tree," a name that suggests it was once used to draw berry borer away from the main crop, and that on many estates its cherries are left to the pickers because harvesting the tall trees is not worth the owner's labour cost.
The S795 connection: Liberica is already in your cup
Liberica's lasting contribution to Indian coffee came through breeding.
The species carries SH3, a gene that confers resistance to coffee leaf rust. At the Coffee Experimental Station in Balehonnur, a selection known as S288 incorporated Liberica-derived material. Under R. L. Narasimhaswamy, S288 was crossed with Kent, a rust-tolerant arabica selection from Mysore. The result was S795, released in the mid-1940s.
S795 went on to become the most widely planted arabica in India, on an estimated 25–30% of arabica acreage, and it is grown across much of South and Southeast Asia. In ICB's catalog, 158 coffees list it as a primary or contributing variety, more than any other. The full lineage is covered in ICB's guide to S795 and India's other arabica varietals.
So a large share of Indian arabica already carries Liberica ancestry. The species never worked as a replacement crop, but it did work as a parent line.
That success has limits. S795's rust resistance has weakened as new races of the fungus have emerged, which is one reason estates have moved toward newer selections such as Chandragiri. It is also why the species behind the original resistance are getting a second look.
Why Liberica is rare in India but Excelsa is spreading
The two species reached India together and were abandoned together. Since then their paths have split, and elevation explains most of the difference.
Liberica's typical cultivated range is very low: roughly 10 to 100 metres, in warm, wet climates with little dry season. Its main producers today, in lowland Malaysia, the Philippines and Indonesia, match that profile. Indian specialty coffee is grown far higher. The estates behind the Liberica and Excelsa lots in ICB's catalog are listed between 850 and 1,200 metres: Mooleh Manay at 850–1,000 m, Dimbada in Sakleshpur at 950 m, Salawara in Sakleshpur at 1,143 m, and Bella Bean India at 1,200 m.
Liberica does grow at these elevations, as the Salawara lot shows, but it grows well outside the range it is known to be most productive in. Add the tree height, the uneven ripening and a large cherry with a high share of pulp to bean, and the economics of harvesting it remain difficult.
Excelsa's typical range is 500 to 1,500 metres, which covers most Indian coffee land. It copes with higher temperatures and longer dry seasons than Robusta. The catalog already reflects the fit: 15 Excelsa lots against 5 Indian-grown Liberica lots, from estates sitting inside Excelsa's range. Its thinner pulp and parchment give a better outturn and make processing easier. SICC's species profile lists it as high-yielding with low nutrient needs, tolerant of leaf rust, resistant to nematodes and resistant to white stem borer, though susceptible to shot hole borer.
White stem borer is the part of that list with the most weight in India. The beetle is a chronic pest of arabica in Karnataka, where larvae bore into the main stem and kill the plant. A species that resists it and fruits at estate elevations gives growers a practical option to test. No published Indian field trials were found for this guide, so the claim rests on species-level characterisation rather than local data. Excelsa cultivation is growing in India, Uganda, South Sudan, Vietnam and Indonesia.
What the Indian catalog actually holds
ICB's catalog data, queried on 26 September 2026, shows how small and concentrated the domestic supply is. Twenty-one lots is a small sample, so the figures below describe what is on sale now rather than a long-run pattern.
Of 1,499 active coffees, 15 are Excelsa (1.0%) and 6 are Liberica (0.4%), against 1,033 Arabica and 58 Robusta. Seventeen roasters sell at least one of the two species; two of them sell both.
Liberica comes mostly from one estate. One of the six Liberica lots is imported: Bermito Coffee Roasters sells a Malaysian lot from the MY Liberica farm in Johor. That leaves five Indian-grown Liberica lots. Four of them come from Salawara Estate in Sakleshpur, sold by Bloom Coffee Roasters, Hill Groove Coffee, KOHI Roasters and Rossette Coffee. The fifth is from Mercara Gold Estate in Coorg, sold by the estate directly.
Excelsa comes mostly from one supply chain. Nine of the 15 Excelsa lots trace to Mooleh Manay Estate or to SICC community lots, sold by Blue Tokai Coffee (two lots), Coffeeverse, Hill Groove Coffee, Maverick and Farmer Coffee, Naivo Coffee, Roastery Coffee, Siolim Coffee and Rossette Coffee. Roaster listings spell the estate at least three ways (Mooleh Manay, Mooley Manay, Mooleh Maney), and ICB groups them under one estate record. Two more lots come from Dimbada Estate in Sakleshpur (both from Genetics Coffee), one from Bella Bean India (Bean By Nation), and three list no estate at all.
With nine roasters selling coffee from one estate's trees, most of the difference between those bags comes from fermentation and roasting choices rather than origin.
Most lots are fermentation lots. Of the 21, fourteen are anaerobic, carbonic maceration or labelled experimental. Five are natural, one is labelled "other," and one is washed, an Excelsa from Ainmane. By species:
- Excelsa (15): 5 carbonic maceration, 4 natural, 3 anaerobic, 2 experimental, 1 washed
- Liberica (6): 4 anaerobic, 1 natural, 1 other
For anyone trying to learn what the species tastes like, this is the complication. A buyer meeting Indian Excelsa is meeting a fermentation style at least as much as a species. The techniques themselves are covered in ICB's guide to anaerobic and experimental processing in India.
Measured on the cheapest listed pack, Excelsa's median price is ₹488 per 100 g (range ₹124–800, 15 lots). Liberica's is ₹320 per 100 g across the five lots with pack weights. For comparison, the Arabica median in the catalog is ₹300 per 100 g and Robusta's is ₹210. Excelsa carries a clear premium. Liberica, on a very small sample, sits close to Arabica. Excelsa is listed as a high-yielding species with easy processing, so its retail premium tracks scarcity and fermentation work more than cost at the farm.
Six of the 21 lots are flagged as limited releases: three Liberica and three Excelsa.
Info: An open question. Coffee Consulate, a German importer, reports buying about 500 kg of Liberica a year from two Indian estates, Badra Estates and Palthope Estate, in both natural and fully washed preparations, and says Palthope is planting seven hectares of new Liberica. Neither estate's Liberica appears in ICB's domestic catalog. It fits a pattern covered in ICB's piece on export versus domestic specialty coffee: some of India's most distinctive lots leave the country before Indian roasters list them.
What they taste like, and how to approach a bag
Descriptions of both species vary widely, and the processing on Indian lots accounts for a good part of that variation.
SICC's species profile describes Liberica as built on tropical fruit (jackfruit, mango, passion fruit, longan) alongside green-tea notes, low acidity and a honeyed sweetness. Bloom's listing for its Salawara natural lists dried fruit, berry, almond and caramel, with more body than a typical arabica. Mercara Gold describes its own Liberica as "richly acidy," which runs against the low-acidity description. Both can be accurate for the lots in question.
For Excelsa, SICC's profile lists dark and dried fruit (currant, prune, plum, fig) with chocolate and spice, low to medium acidity, low bitterness and a rounded sweetness. Genetics Coffee's Dimbada Excelsa lists dried figs, prunes and dried berries, which sits close to that description even after a cashew-fruit fermentation.
These descriptors work like any tasting note: they indicate the direction of a cup rather than promise a result.
Tip: Because most Indian lots are fermented, the ferment accounts for part of what reaches the cup. Tasting a natural or washed lot, such as Bloom's Salawara natural Liberica or Ainmane's washed Excelsa, next to a carbonic or anaerobic one is the clearest way to separate the species from the process. No special brew method is needed; these behave like other light-to-medium specialty roasts.
Who these coffees are for
Drinkers already exploring fermented, fruit-forward coffees will find most of the Indian supply in familiar territory. The current lots are mostly experimental-process coffees that happen to use an unusual species.
Drinkers interested in estate genetics and climate adaptation have a different reason to look. These are the species that Indian breeding drew rust resistance from, and Excelsa's field profile makes it a candidate for estates dealing with heat, dry spells and white stem borer.
For either group, the supply pattern is the same: a handful of estates, lots tied to single harvests, and a premium on Excelsa over the Arabica median. Lots appear and disappear with each season. The current listings are on the Excelsa and Liberica discovery pages.
References
- Davis, A. P. et al. (2025). Genomic data define species delimitation in Liberica coffee with implications for crop development and conservation. Nature Plants 11: 1729–1738.
- Davis, A. P. et al. (2026). Genomic elucidation of hybridization between Liberica and excelsa coffee and its implications for coffee crop development. Scientific Reports.
- Davis, A. P. et al. (2019). High extinction risk for wild coffee species and implications for coffee sector sustainability. *Science Advances* 5(1): eaav3473
- Hull, E. C. P. (1877). Coffee Planting in Southern India and Ceylon. London: E. & F. N. Spon.
- Government of Mysore. Report of Work on the Coffee Experimental Station, Balehonnur, 1925–1930.
- Narasimhaswamy, R. L. (1960). Arabica Selection S.795: Its Origin and Performance. Coffee Board of India.