Marsellesa

Marsellesa is a modern rust-resistant coffee variety, released in 2009 in Nicaragua by CIRAD–ECOM, a partnership between a French public agricultural research institute and one of the world’s largest green-coffee traders. Its parents are Timor Hybrid 832/2 × Villa Sarchi CIFC 971/10, which makes it a member of the Sarchimor family.¹

In plain terms: someone took a compact Costa Rican dwarf coffee tree that tastes good and gets sick, crossed it with a natural Arabica–Robusta accident that tastes worse and doesn’t get sick, and spent years selecting the offspring for the best of both. That is the whole Sarchimor idea, and Marsellesa is the newest and most commercially deployed version of it.

Two things make it worth its own note rather than a row in the Sarchimor table. It is the highest-yielding line in that family that still rates “Good” in the cup. And it is the mother of Starmaya — a first-generation hybrid that broke a long-standing practical barrier in coffee breeding.

What the catalogue actually says

Every figure below is from the World Coffee Research variety record for Marsellesa, confirmed directly at source.¹

Field	Marsellesa
Genetic description	Introgressed (Sarchimor related)
Lineage	Timor Hybrid 832/2 × Villa Sarchi CIFC 971/10
Breeder	CIRAD–ECOM
Country / year released	Nicaragua, 2009
Stature	Dwarf / compact
Leaf tip colour	Green
Bean size	Average
Yield potential	High
Quality potential at high altitude	Good
Coffee leaf rust	Intermediate resistance
Coffee berry disease	Tolerant
Nematodes	Susceptible
Nutrition requirement	High
Ripening	Average
Planting density	5,000–6,000 plants/ha, single-stem pruning

The optimal altitude is given as a sliding band, because the same tree wants different elevations depending on how far it sits from the equator:¹

	Latitude	Optimal altitude
1	5°N–5°S	1,000–1,600 m
2	5–15°N and 5–15°S	700–1,300 m
3	Above 15°N / 15°S	400–1,000 m

The catalogue describes it as “adapted to medium altitudes” and as “a pedigree selection made by ECOM-CIRAD in Nicaragua specifically for its rust resistance traits.”¹

Two practical readings. “Nutrition requirement: high” plus 5,000–6,000 trees per hectare means this is an expensive tree to farm properly — it rewards farms with money for fertiliser and punishes farms without. And “susceptible” to nematodes, the microscopic root worms that stunt coffee trees, is a real gap: a variety bred to survive one disease has no defence against a different one.

To verify

The rust rating is routinely overstated in the trade. Multiple seller and glossary pages describe Marsellesa as having “strong resistance to coffee leaf rust,” and the research report behind this note attributed that phrasing to World Coffee Research itself.² The catalogue does not say that. Its rating for Marsellesa is “Intermediate” — the same rating given to T5296 and IAPAR 59, and a step below Parainema, the only Sarchimor rated “highly resistant.”¹³ This wiki’s Sarchimor note already records Marsellesa as intermediate. Treat “strong rust resistance” as marketing copy, not a catalogue fact.

Is it “T5296”?

Probably not, and this one is worth untangling because the claim circulates.

Sweet Maria’s glossary says Marsellesa “also goes by the interesting name of T5296.”⁴ But T5296 has its own separate record in the World Coffee Research catalogue, with a different breeder (CIFC, trialled at CATIE), a different country (Costa Rica), a different date (1971 field trials) and a different yield rating (medium, not high).³ Two catalogue entries with different breeders and a 38-year gap are not the same variety.

The likeliest explanation is descent rather than identity — both are Villa Sarchi × Timor Hybrid 832/2 crosses, and Sarchimor breeding material moved between institutes. But that is an inference.

To verify

“Marsellesa = T5296” is a single-source claim that contradicts the catalogue. It appears on Sweet Maria’s glossary page⁴ and is not repeated in World Coffee Research or CIRAD material located here.¹ The two varieties hold separate catalogue entries with different breeders, countries, release dates and yield ratings.¹³ Do not use the two names interchangeably.

What it tastes like

The catalogue’s own cup note is short: “notably high acidity in the cup.”¹ Its formal rating is “Good” quality potential at high altitude — the same rating Caturra gets, and the ceiling for anything in the Sarchimor or Catimor families short of Parainema.¹

That is the entire body of independent quality evidence for this variety. Everything else in circulation is a seller describing coffee it is trying to sell.

To verify

No Cup of Excellence result for Marsellesa was found. A green-coffee seller’s glossary states that Marsellesa’s quality has been “sufficient to appear in Cup of Excellence results,” but names no year, country, farm or lot.² The Alliance for Coffee Excellence results pages for Nicaragua 2024 and Nicaragua 2025 — the variety’s home country, its two most recent editions — list no Marsellesa or Sarchimor lot in either year.⁵⁶ (Parainema does appear in both.) This does not disprove the claim for other countries or earlier years, but nothing supports it either. Treat any “Marsellesa has placed in COE” statement as unverified.

To verify

The farm-level lots and their scores are self-reported by sellers. Four Marsellesa offerings surfaced in the research behind this note, each on a single roaster or importer page, none independently corroborated, and none of the scores tied to an SCA-certified session or a competition:

• Volcan Azul, Alajuela, Central Valley, Costa Rica — producer Alejo Castro, 1,500 m, yellow honey anaerobic, harvested January–March 2021, 600 kg bought by April Coffee Roasters; notes of melon and nougat. No score given — do not infer one.⁷
• Acatenango Valley, Guatemala — producer Brandon Estrada, 1,550 m, natural, SHB grade, roasted on-farm; claimed 85 points; notes of chocolate, citrus, nuts, caramel.⁸
• Chimaltenango, Guatemala — producer Roberto Matheu, farm established 1865, ~1,829 m (6,000 ft), washed and patio-dried; claimed “85+” with no exact figure or scoring body.⁹
• Mount Sunzu, near Mbala, northern Zambia — natural, claimed 84.5; altitude not stated on the listing.¹⁰

Separately, a Sucafina listing sells a Hacienda Sonora lot in Costa Rica’s Central Valley as “Marsellesa,”¹¹ while this wiki’s Sarchimor note records the same farm’s lots listed elsewhere under the plain family name “Sarchimor.” Different blocks or looser labelling — not resolved here.

Its most important offspring

Marsellesa’s biggest contribution may not be as a farm tree at all. It is the pollen parent of Starmaya.¹²¹³

Here is why that matters. F1 hybrids — the first generation of a cross between two genetically distant parents — are the star performers of plant breeding: they tend to out-yield and out-cup both parents. The catch in coffee is that F1 seed does not breed true, so growers historically had to buy clones, propagated one plant at a time in a lab. That is expensive enough to keep F1 hybrids out of reach for most farmers.

In 2001, CIRAD researchers found a naturally male-sterile Ethiopian coffee plant — a tree that produces no pollen of its own — while screening accessions at the CATIE germplasm bank in Costa Rica. It was named CIR-SM01.¹³ Plant it beside Marsellesa and wind or insects do the crossing for you: every cherry the sterile mother sets is necessarily a hybrid. The seed garden that does this was planted in 2009 at La Cumplida farm in Nicaragua.¹³ That cross is Starmaya, and World Coffee Research calls it “the only F1 hybrid in the world that professional seed producers can propagate by seed.”¹²

(An earlier version of this note put the 2001 discovery at La Cumplida and called Marsellesa the mother plant. Both were wrong: CIR-SM01 is the seed-bearing mother, Marsellesa is the pollen donor, and La Cumplida is the 2009 seed-garden site.¹³ See Starmaya.)

Starmaya rates highly resistant to Coffee leaf rust and “very good” in the cup — both a step above Marsellesa’s own ratings.¹² The parent that cups “good” produced a child that cups “very good,” and out-yielded Marsellesa itself by about 30% in Nicaraguan trials.¹³

Where it is grown

The catalogue places Marsellesa in Nicaragua.¹ The farm listings above put commercial lots in Costa Rica, Guatemala and Zambia, and glossary sources add Honduras.⁴²

To verify

Reported spread beyond Central America rests on single sources. A varieties blog reports “trials and commercial plantings in Laos and Vietnam” with good rust performance, and gives an optimal altitude of 900–1,600 m — a band that does not match the catalogue’s latitude-adjusted figures above.¹⁴ Sweet Maria’s reports plantings “throughout Central and South America” on the basis of farm visits and photographs.⁴ Neither is corroborated by a breeder or institute document located here.

Why it matters

Catimor is the reason “rust-resistant hybrid” became an insult in specialty coffee — the early lines survived the disease and lost the cup. Marsellesa is what the same bargain looks like twenty years of breeding later: same rust genes, same dwarf frame, but high yield and good cup instead of low. It is not an exotic; the catalogue’s own summary is one line about acidity. What it is, is the credible middle — a tree a Nicaraguan or Honduran farm can plant without either losing the crop to rust or being written off by buyers.

And through Starmaya it did something the heirlooms could not: it made a high-performing hybrid cheap enough to hand out as seed.

Related

• Sarchimor — the family it belongs to, and where it sits against the other three named selections
• Villa Sarchi — the dwarf Bourbon mutation that is its Arabica parent
• Timor Hybrid — the Arabica × Robusta parent that carries the rust resistance
• Robusta — the second species behind that parent, and the original source of the resistance genes
• Starmaya — its offspring, and the only seed-propagated F1 hybrid in coffee
• Parainema · IAPAR 59 — the Honduran and Brazilian Sarchimor selections
• Catimor — the sibling family built on Caturra, and the cautionary tale Marsellesa was bred against
• Coffee leaf rust — the disease the whole family exists to survive
• CIRAD · ECOM — the research institute and the trading house that bred it
• CIFC — keeper of the accession numbers in its pedigree
• World Coffee Research — publisher of the variety catalogue every figure here comes from
• Cup of Excellence · Alliance for Coffee Excellence — the competition whose results were checked above
• Cupping · Coffee scoring · SCA cupping protocol — how the quality claims above would be settled
• Natural process · Honey process · Washed process · Anaerobic fermentation — the processing styles on the farm lots listed above

Official links

• Official site — World Coffee Research variety record: Marsellesa
• Official site — World Coffee Research variety catalogue
• Official site — CIRAD
• Official site — ECOM Agroindustrial Corp.

Footnotes

1. World Coffee Research — Marsellesa ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰

2. Green Coffee Collective — Marsellesa glossary entry ↩ ↩² ↩³

3. World Coffee Research — T5296 (Sarchimor) ↩ ↩² ↩³

4. Sweet Maria’s Coffee Library — Marsellesa glossary entry ↩ ↩² ↩³ ↩⁴

5. Alliance for Coffee Excellence — Cup of Excellence Nicaragua 2025 results ↩

6. Alliance for Coffee Excellence — Cup of Excellence Nicaragua 2024 results ↩

7. April Coffee Roasters — Marsellesa lot information ↩

8. El Conecte Coffee — Marsellesa Farmlevel Reserve ↩

9. Glen Ellyn Coffee Co. — Marsellesa ↩

10. Green Coffee Collective — Natural Marsellesa AB Zambia ↩

11. Sucafina — Hacienda Sonora Natural Marsellesa ↩

12. World Coffee Research — Starmaya ↩ ↩² ↩³

13. Georget et al. (2019) — “Starmaya: The First Arabica F1 Coffee Hybrid Produced Using Genetic Male Sterility,” Frontiers in Plant Science 10:1344 ↩ ↩² ↩³ ↩⁴ ↩⁵

14. Volcana Coffee — Marsellesa coffee variety ↩