Sri Lanka

Why this origin matters

Sri Lanka matters to anyone learning about coffee because of what happened there between 1869 and 1886. As Ceylon it was, before the epidemic, the world's third largest producer on one scientific review's account. Coffee leaf rust was first recorded as an epidemic there, and within two decades an export trade that had averaged more than nine hundred thousand hundredweight a year had fallen by four-fifths and the estates were turning to tea. The episode is the reason the fungus was once called the Ceylon leaf disease, and a scientific review calls it one of the best-known in the history of plant pathology. It is also a history of how a plantation economy was made: on land taken into estates in the Kandyan highlands and with labour recruited from South India and held by debt. Today's crop is small, 5,117 tonnes of green coffee in 2024 on the figure the country reported to the FAO, and the documents CoffeeHQ could open say little about it.

Where coffee grows

The Department of Export Agriculture names two sets of districts. It describes arabica as mostly grown in Nuwara Eliya, Kandy, Matale and Badulla, in the island's wet and intermediate zones, and robusta as mainly grown in Kegalle, Kurunegala, Kandy and Matale. Its page gives no area or production for any district, and the total extent it prints is not credible and is not used here. The nineteenth-century plantations were in the same highlands: a historian describes them as opened in the Kandyan region after the British conquered the kingdom of Kandy in 1815 and built a road to it from Colombo, and says the Dutch before them had failed at large-scale planting for want of suitable land in the middle and upper country. A 1985 Sri Lankan review adds that the Dutch had popularised coffee in the coastal districts of Galle and Negombo. The epidemic itself is placed by two sources at Madulsima, where the rust was first noticed in 1869.

The only present-day altitude in the documents read is a recommendation, not a description of where the crop stands. The Department of Export Agriculture advises growers that elevations above 800 metres suit arabica and elevations up to 800 metres suit robusta, and that the hybrid it spells Catimur can be grown at both. It gives a mean annual temperature of 18 to 24 degrees Celsius and rainfall of 1,500 to 2,750 millimetres for arabica. A historian of the early plantations notes that the first British ventures of the 1820s failed partly because their land lay too low.

Species and varieties

Two species are grown commercially, on the department's account: arabica and robusta, which it identifies as Coffea canephora. It says Liberica is not grown commercially in Sri Lanka. The arabica varieties it names as high-yielding are, as its page spells them, HTD, S09, Catimur, Lakparakum, Lak Saviru and Lak Komali, of which it says some are resistant to rust; the robustas it lists as newly released are IMY, S274, GCR, CCI, Lanka Chandra, Lanka Bimsara and Lanka Isuru. The page gives no parentage for any of them and CoffeeHQ has found none. Catimur is presumably Catimor, the page calling it a hybrid; that is CoffeeHQ's reading of a spelling and not a statement in the document, and nothing is said here about what the other names are. It does say that the return of coffee followed the introduction of rust-tolerant arabicas and hybrids. A century earlier the experiment stations were already looking beyond arabica: a 1922 trade history has Peradeniya and Anuradhapura testing robusta and related coffees in the hope of reviving the industry.

What the export agriculture department tells growers now

The department's page is advice to smallholders, and it describes coffee as one crop among several. It recommends growing it mixed with pepper, clove, nutmeg, coconut and tea, under shade trees such as gliricidia or with bananas for temporary shade, and feeding it, if nothing else is to hand, with the fertiliser mixture sold for tea. Arabica is to be planted at about 3,000 plants a hectare and robusta at 1,150 to 1,600.

It treats rust as a problem of the past that resistant varieties have answered, while still calling arabica highly susceptible, and names the coffee berry borer as a present pest. Its history of the crop is four sentences long and ends with imports: after the collapse, it says, nearly 600 tons of coffee a year were brought into the country, a figure it does not date.

Harvest

No present-day document read names Sri Lanka's harvest months. The one statement found is historical: a study of nineteenth-century plantation labour gives the coffee harvest in Ceylon as October to February, in explaining why workers came and went with the season before tea, which is picked all year, replaced it. The department's page mentions limiting the main harvest to three rounds as a defence against the berry borer, and says no more about timing.

How the coffee is processed

The Department of Export Agriculture describes two methods in use. In the dry method ripe fruit is dried in the sun and then hulled, and the product is called cherry coffee. In the wet method the fruit is pulped within ten hours of picking, fermented under a mat for 24 to 48 hours, washed and dried in the sun, and the product is called parchment coffee, which the department says is the better of the two. It gives no share for either.

How it is graded and sold

Sri Lanka exports little coffee and no recent trade figure was read. Two older points in the record are worth having. A 1922 trade history says Ceylon coffees were classed commercially as native, plantation and mountain, and were sold almost entirely in London; by then, it adds, the island no longer grew enough for its own use and ranked as an importer. And a 1985 Sri Lankan review records a second, much smaller export trade: 132,100 kilograms in 1973, rising to a record 3,124,400 kilograms in 1983, which it connects with the country's inclusion in the International Coffee Agreement and the quota that came with it. The same review estimated the area under coffee in 1985 at well over 30,000 acres, without saying how it was counted.

History

Coffee was in Ceylon before the plantations. A 1922 trade history repeats, as legend, that Arab traders brought it before the Portuguese arrived in 1505, and dates Dutch cultivation from 1658 and more systematic planting from 1690; the Department of Export Agriculture gives 1503 for the Arab introduction, without a source. A historian of the period says the Kandyan peasantry were already growing small amounts, and that most of the coffee exported up to the early 1830s, around 10,000 hundredweight a year, was theirs. Estates came with British rule. Land near Gampola was granted to a colonel in 1823 and 1824 and the governor, Edward Barnes, planted coffee himself, though most of these first ventures failed. What changed in the 1830s, on the same historian's account, was the market: Britain abolished slavery in its West Indian colonies, equalised the duty on West Indian and Ceylonese coffee, and so ended the protection its Caribbean planters had enjoyed. Capital and planters poured in, and during what was called the coffee mania of the 1840s the island was remade as a plantation economy with coffee as its single crop for the next forty years. The trade history records the other side of that decade, a speculation said to have involved five million pounds sterling that ended in a crash in 1845.

Labour recruited in South India, and held by debt

The estates were worked almost entirely by wage labourers brought from South India, first for coffee and then for tea. A historian of the system, Roland Wenzlhuemer, describes where they came from: districts of the Madras Presidency where landlessness was widespread, droughts and famines frequent, and where British manufactured goods had put village artisans out of work. Sinhalese villagers also worked on the early estates, and how much is debated among historians; but planters with what he calls a plentiful, willing and cheap supply at hand did not try hard to recruit them.

This migration was called free, because it was not the indenture under which Indian labourers were shipped to other colonies from 1834. Historians put the word in inverted commas, and the study explains why. Recruiting was done by kanganies, at first workers who brought kin and neighbours and, from the late 1840s, professional recruiters paid a premium for every labourer delivered. Planters advanced them money for the journey, and each labourer's share of that advance was deducted from wages, so that a worker arrived on the estate already in debt, often for sums never spent on him. Contemporaries estimated that recruiters spent less than a third of the advance on the journey and that hundreds of migrants died of starvation on the way; the study also records kidnapping. Wages and the monthly rice ration passed through the kangany's hands.

The law stood on the planter's side. Under the Master and Servant ordinances of 1841 and 1865 a labourer could be fined or imprisoned for up to three months for neglect of duty, for a wide range of misconduct or for leaving without a month's notice; the only offence an employer could commit was refusing to pay wages. In theory a labourer could leave on a month's notice. In practice, the study concludes, debt and the kangany's hold kept most from doing so, and the colonial government took almost no part in regulating the migration until late in the century. Patrick Peebles, a historian with whom the study argues over how seasonal the work was, gives these workers their usual name in the title of his book, The Plantation Tamils of Ceylon.

How fast it fell: 939,000 hundredweight to 179,000 in under twenty years

The historian's table, built from colonial figures he warns are incomplete, gives annual averages. Exports averaged 939,000 hundredweight a year in 1865 to 1869, the highest period in it. They were 881,000 in 1870 to 1874, 795,000 in 1875 to 1879 and 433,000 in 1880 to 1884, then 316,000 in 1885 and 179,000 in 1886.

The other columns show what the fall looked like from the estates. Through the 1870s the area planted went on growing, from an average of 243,000 acres in the late 1860s to 310,000 in the late 1870s, and the price roughly doubled, from 52 shillings a hundredweight to 108. Planters were opening new land and being paid more for less. A Sri Lankan review of 1985 says the same in words: yields were dwindling while the acreage grew, and the damage was not felt at first because prices were booming. That is the review's explanation, and the table is consistent with it; neither is a study of what planters knew.

Tea filled the space quickly. The same study reprints the colony's figures: 14,266 acres under tea in 1880 and 235,794 in 1890, with exports rising from about 163,000 pounds to 45.8 million. The trade history has the coffee tree practically gone from the island by 1887, and lists what was left afterwards: 4,875 acres in 1903 and under a thousand by 1919.

The sources do not agree on the peak, and CoffeeHQ does not choose between them. The 1922 trade history gives the record export year as 1873, at 111.5 million pounds, in one chapter and as 1870, at 118.2 million, in another. For the largest area under coffee it says 340,000 acres at one time; the 1985 review says 275,000 acres in 1878; the historian's five-year average for 1875 to 1879 is 310,000.

The rust, and what two sources name beside it

The fungus is Hemileia vastatrix, described and named in 1869. Two sources place its first notice in Ceylon that year at Madulsima. A scientific review says it wiped out coffee cultivation on the island soon after, with devastating social and economic consequences, and the 1922 trade history says flatly that the disease ruined the industry. The labour historian says the same: a leaf fungus virtually wiped out the plantations in the late 1870s and early 1880s, and the planting enterprise was practically dead by 1885.

Two of the sources read add something to the fungus. A 2021 review in the Proceedings of the National Academy of Sciences says that by the 1890s nine coffee farms in ten in Ceylon had been abandoned, and that this was triggered in part by the fall in production from the rust but also by a fall in coffee prices in the early 1880s, which saw many farmers move to cinchona, tea and other crops. It says so in one sentence, citing historical work CoffeeHQ could not open. The 1985 Sri Lankan review, which itself calls the blight the fatal blow, describes it coinciding with the depression of 1879 to 1884, when it says demand and prices dropped considerably, and with bank failures, among them the Oriental Bank, that cut off the credit a recovery would have needed. The labour historian's own table does not show prices falling below where they stood before the disease: the unit value of exports was 52 shillings a hundredweight in the late 1860s, 108 in the late 1870s, 89 in 1880 to 1884 and 78 in 1885. That is a fall from a peak, on a much smaller crop. So this entry can report that prices and credit are named beside the rust, and cannot say how much weight they carry. The historian, for his part, calls the collapse a prototypical example of the hazards of monoculture, and credits cinchona and then tea with ending the financial crisis that followed.

Why the disease was so severe there is a separate question, and here the reviews agree with one another. The arabica carried out of Yemen and planted across Asia had never been selected in the presence of rust, so it had no resistance to lose. Ceylon grew it in large, uniform plantations in full sun, and one of the reviews, citing a paper of 1882 by Harry Marshall Ward, says a direct link between that intensification and the severity of the disease was reported during the epidemic. The epidemic was, in one review's words, the outcome of genetic, biological and agronomic circumstances. Rust then spread through the coffee lands of the Indian Ocean and the Pacific between 1870 and 1920. The Indonesia entry and the Timor Hybrid page take the story on from there.

Pressures on the sector

• A thin record. No present-day document read gives the area, the number of growers, the harvest months or the trade, so the size and shape of today's sector cannot be described.
• Rust and the berry borer, both named by the Department of Export Agriculture as affecting the crop, with arabica described as highly susceptible to rust.
• Coffee as a minor crop grown mixed with spices, coconut and tea, on the department's own advice, which makes it hard to count.

Institutions worth knowing

• Department of Export Agriculture — the government department whose crop page gives today's recommended districts, elevations and varieties
• The experiment stations at Peradeniya and Anuradhapura — named in a 1922 trade history as then testing robusta and related coffees to revive the crop

Explore Sri Lanka's coffee from here

A reading order, not a list of everything related: each step picks up something this page raised and takes it further.

1. Step 1Coffee pests and diseasesWhat leaf rust is and how it kills a crop, which is the disease first recorded as an epidemic in Ceylon.
2. Step 2Timor HybridThe natural cross that gave breeders a source of rust resistance, a century after the epidemic in Ceylon.
3. Step 3CatimorThe rust-resistant group that the hybrid on the department's list appears, from its spelling, to belong to.
4. Step 4IndonesiaWhere the same rust arrived next, and how an arabica colony became a robusta country.
5. Step 5RobustaThe species Ceylon's experiment stations were already testing by 1922, and one of the two the department says Sri Lanka grows commercially now.
6. Step 6BrazilAnother coffee economy built on coerced and recruited labour, set out with the same attention to who did the work.
7. Step 7Washed processingThe wet method the department describes, step by step, with what fermentation is for.