Coffea arabica L.

Where it comes from

Not Arabia. Kew gives the native range as eastern South Sudan, south-western Ethiopia and northern Kenya — Mount Marsabit specifically.¹ Yemen and the rest of the Arabian peninsula are on the introduced list, and NC State calls the epithet a misnomer outright.²

It grows as an understorey tree in montane forest between roughly 950 and 1,950 m.³ That single fact settles most of the care questions: it wants dappled light rather than a sunny windowsill, it wants a mean temperature around 64-73 °F (18-23 °C), and it is not a plant for a hot bright room.

It is Endangered in the wild.³⁴ Which is worth sitting with for a moment, given how much of it the world drinks.

Naming and description

Linnaeus published the name in Species Plantarum in 1753.¹⁵ Kew accepts 133 species in the genus,⁶ and the former genus Psilanthus is now folded into it, which is why the genus range reaches as far as Queensland.⁷

Fun facts

• “Three to four years to fruit” is a plantation figure that lost its label. It comes from the commercial-cultivation paragraph of an extension page — the sentence sits between one about growing for beans and one about pruning plantations to six feet for harvest.² The same page’s indoor paragraph gives no timescale at all, and nothing I could reach documents years-to-flowering for a plant in a pot. The hobby did not misread field agronomy; it copied it accurately and dropped the context.
• The physiology is more useful than a number, and it explains why yours has not flowered. Bud initiation needs three things an ordinary room defeats: a photoperiod of 13 hours or less, temperatures below 82 °F (28 °C), and a dry period followed by re-wetting to break floral bud dormancy.⁸ Kew independently lists a defined dry season among its requirements.³ A plant kept warm, evenly lit and evenly watered is being told three separate ways not to flower. And even when it does, it is 220-243 days from flower to ripe cherry⁸ — seven or eight months, not weeks.
• The species is a natural hybrid, and a recent one. C. arabica is an allotetraploid formed from C. canephora and C. eugenioides. Genome work dates the founding polyploidy event to somewhere between 350,000 and 610,000 years ago,⁹ though an earlier study declined to pin it at all, citing a published range spanning 10,000 to 665,000 years.¹⁰ Either way it is young, and its famously narrow genetic base follows from it.
• Caffeine is a defence compound that also manipulates pollinators, by the same mechanism that poisons a dog. Caffeine in nectar made bees three times as likely to remember a scent, at doses below their own bitter-taste threshold — the plant is buying loyalty.¹¹ It works by blocking adenosine receptors. In a dog the same blockade, at a much higher dose, gives 20 mg/kg for mild signs, 40-50 for cardiotoxicity and 60 upwards for seizures.¹²
• The ASPCA page called “Coffee Tree” is not coffee. Its scientific-name field reads Polyscias guilfoylei — an aralia, in a different family entirely, with saponin as the toxic principle.¹³ There is no ASPCA record for Coffea arabica at all. This is the fifth consecutive group on this site where an ASPCA page title has named a different plant from its own scientific-name field.
• A roasted bean will never grow. Coffee seed has intermediate storage behaviour — it does not survive drying — which is why it must be sown fresh and why the packet in your cupboard is dead.¹⁴