CoffeariumBeta
  1. Plant
  2. Farm
  3. Process
  4. Trade
  5. Roast
  6. Brew
  7. Cup
Botany · Fruit and seed

Flowering and pollination

How the coffee plant forms flower buds, why they wait for rain after a dry spell, how long a flower lasts, and how arabica pollinates itself while canephora needs pollen from another plant – and what bees add.

Across the chain: Science

Source-checkedby Coffearium verification pass· Last source-checked 22 Oct 2026· Prepared with AI assistance· 14 sources (13 tier A)· 3 contested· Updated 22 Oct 2026
Reading depth
Contents
  • Coffee flower buds stop growing when mature and rest, often for 30–120 days, until rain or irrigation after a dry spell wakes them.1,2
  • Flowers open about 1–2 weeks after the waking rain; reported intervals range from 7 to 14 days, depending on species and study.4,5
  • A flower stays open for 2–3 days; pollen stays viable for 24–36 hours and the stigma is receptive for 48 hours.1
  • *C. arabica* mainly pollinates itself; in Colombia, insect cross-pollination accounted for 9.4–26.7% of fruit set.9,1
  • *C. canephora* is self-incompatible: one gene with several alleles decides which plants can pollinate each other.11

From bud to open flower#

A coffee plant does not flower at any time. Flower buds form in the leaf axils and grow for weeks. When they are mature, they stop growing and rest.1 This resting phase is called dormancy. It usually lasts 30–120 days.1

Short days start the process. Coffee is a short-day plant: more than 12 hours of daylight delay the start of flower buds.1 Heat also blocks it. One review gives 28 °C as the limit; another cites over 30 °C by day.1,4

The resting buds open after water returns. Under natural conditions, the first rains after a dry period often break their dormancy.2 Irrigation after a dry spell does the same.3,5 The buds then grow quickly and open. This opening is called anthesis.1

How long from rain to bloom#

The open flowers appear about one to two weeks after the rain. Reported intervals differ by species, by the size of the flowering and by study.4,5 For C. arabica, a 2025 review gives 10–14 days; for C. canephora 8–11 days.4 In a greenhouse experiment in Hawaii, arabica buds needed 9–12 days after re-irrigation.5

0246810121416daysC. arabica, re-irrigationafter water deficit(Hawaii, greenhouse)9–12 days5 C. arabica (2025 review)10–14 days4 C. arabica, largeflowering after rain of 10mm or more(Nicaragua)8–10 days4 C. arabica, smallerflowering(Nicaragua)10–13 days4 C. canephora (2025 review)8–11 days4 C. liberica (2025 review)7–10 days4
Data table
daysSource
C. arabica, re-irrigation after water deficit (Hawaii, greenhouse)9–12 days5
C. arabica (2025 review)10–14 days4
C. arabica, large flowering after rain of 10 mm or more (Nicaragua)8–10 days4
C. arabica, smaller flowering (Nicaragua)10–13 days4
C. canephora (2025 review)8–11 days4
C. liberica (2025 review)7–10 days4
Days from rain or re-irrigation to open flowers, as reported in a 1992 experiment in Hawaii and a 2025 systematic review. Sites, varieties and how the start was defined differ between rows.4,5

In C. arabica in Nicaragua, a 2025 systematic review reports, large flowerings tended to come 8–10 days after a rain of 10 mm or more, and smaller flowerings after 10–13 days.4 A longer dry spell lets more buds reach the stage that responds, so they open together.3 How the fruit then grows is covered in Fruit development.

One flowering or several#

How many times a plant flowers depends on how rain and dry weather are spread over the year. Near the equator, rain falls in most months and coffee flowers many times.1,3 Farther away, dry seasons are longer and one main flowering is usual.3

Colombia shows this across its coffee regions. Between 4° and 5° N, about 60% of the flowers open from November to April and 40% from May to October.6 Above 10° N, flowers open only from November to April.6 Several flowerings a year lead to uneven ripening and several harvests.6 Harvest seasons by origin are covered in Harvest calendars.

Scattered light showers in the dry season do the opposite of a clear trigger. According to a 2017 textbook chapter, they cause several blossom periods, flower atrophy and fruit that ripens unevenly.8

The flower#

12673458
  1. Corolla lobe (petal): five white lobes, 8.0–14.5 mm long; the corolla is 18.0–30.9 mm across.1
  2. Corolla tube: the lobes open from the top of a narrow tube.1
  3. Anther: a long anther (6.00–9.05 mm) on a short filament (4.1–5.0 mm).1
  4. Style: long and slender, 12.0–15.0 mm.1
  5. Stigma: split in two, 4.3–5.8 mm long; receptive for 48 hours.1
  6. Calyx: the small outer whorl, formed first.3
  7. Ovary: two chambers, one ovule each – the two future seeds.1
  8. Flower bud: after rewetting it gains mass for 6–12 days, then opens within 3–4 days.1
A C. arabica flower seen from the front and slightly above, with a closed bud. Point at a part, its number or its name to highlight it. Schematic, not to scale.

The coffee flower is perfect: it has both male and female parts.1 In arabica, four to five flowers grow together in each cluster.1 The flower has four whorls, formed from the outside in: sepals, petals, stamens and carpels.3 The ovary has two chambers with one ovule each, matching the two seeds of a typical cherry; see Coffee cherry anatomy.1

Each flower lives only briefly. It stays open for 2–3 days, then the corolla and stamens fall off.1 In Indonesia, flowers opened just before dawn and attracted bees for usually only one day.12

Self-pollination and cross-pollination#

C. arabica can fertilise itself. Its mating system is mainly self-fertilisation, with occasional cross-pollination by insects.9 In Colombian field studies, self-pollination gave 56.4–95.3% of fruit set and insect cross-pollination 9.4–26.7%.1

C. canephora cannot fertilise itself. It is self-incompatible: a flower rejects pollen that carries the same compatibility type as itself.11 In one experiment, self-pollinated canephora flowers set 10% fruit or less.10 Farms therefore plant several compatible clones together; see Coffea canephora.11

Bees and other visitors#

Bees visit coffee flowers in both species. Open pollination and hand cross-pollination gave the highest fruit set in arabica and in canephora.10 In Costa Rica, the main visitors were feral honeybees and 10 native species of stingless bees.13

More bee species meant more fruit in one study. In 24 coffee agroforestry plots in Indonesia, fruit set rose from about 60% with three bee species to 90% with 20.12 Shade trees and forest nearby shape which bees visit; see Shade and agroforestry.12,13

Coffee nectar contains caffeine. In a 2013 experiment, honeybees rewarded with caffeine were three times as likely to remember a learned floral scent.14

Floral induction and dormancy#

Flower initiation is controlled by day length and temperature. A 2025 review states that photoperiods above 12 h delay initiation, and that it does not occur above 13 h or above 28 °C.1 A 2025 systematic review cites day temperatures above 30 °C with nights above 17 °C as inhibiting it (see Contested).4 A 2021 review describes the fall in day length below 13 hours after the March equinox as the signal for flower bud development.3 The same review proposes that induction starts in January and extends until July for early and late cultivars, wider than the February–April window given in older literature.3

In Kenya, east of the Great Rift Valley, a 1970 study found main flowerings either in February–March or in October–November. These were mainly initiated from about August to December and June to September, respectively.7 The authors list pruning, tree condition and weather as causes of a switch between the two rhythms.7

Buds pass through BBCH stages 53 and 57 before they are fully differentiated at stage 58.1 At 4–6 mm they stop growing and enter dormancy, according to a 2025 systematic review citing a 2001 phenology study.4 Dormancy at stage 58 is described as true dormancy, maintained by high levels of abscisic acid (ABA); the 2025 review gives 30–120 days, and 90–150 days from initiation to the end of dormancy.1 In thermal time, the 2025 review gives 840 growing degree-days (GDD) from initiation to dormancy, and 1,579 GDD for Brazil.1

Breaking dormancy#

A water deficit followed by rewetting opens dormant buds. A 1992 study on arabica cv. Guatemalan in Hawaii found that re-irrigation opened flowers only if predawn leaf water potential had fallen below −0.8 MPa.5 Milder but longer deficits, about −0.3 to −0.5 MPa for two weeks, had the same effect.5 Only buds at the “open white cluster” stage (stage 4 in that study) responded. Split-root experiments pointed to a root signal, independent of leaf water status.5 A 2007 review cites a threshold water potential of about −1.2 MPa, from a 1976 study, as critical for triggering flower opening, and a 2025 review gives soil water potentials below −1.4 MPa; see Contested.2,1

Rewetting changes the hormone balance between ABA and ethylene.1 About 48 h after rehydration the buds resume growth, then gain fresh and dry mass over 6–12 days depending on temperature; anthesis follows within 3–4 days.1 In thermal time, anthesis needs about 100 GDD after dormancy is broken.1 A longer dry period lets buds at different stages reach the responsive stage, so flowering after rewetting is more uniform.3

The amount of rain needed is not settled. A 2025 review gives 5–10 mm to restart growth and more than 14 mm for adequate flowering.1 A 2025 systematic review gives more than 10 mm in a single day after a dry period averaging under 0.6 mm per day for at least 11 days.4

Flowering regimes#

In equatorial regions without a well-defined dry season, and with an inductive photoperiod all year, coffee flowers many times.6,1 A 2023 Colombian study of 15 arabica accessions at four sites found two patterns: an “annual” one at Pueblo Bello, Cesar (10° N) and a “continual” one in Caldas, Quindío and Cauca (2–5° N).6 Caldas and Quindío had 2.2–4.0 times as many flowering events as Cauca; accessions and tree size made no detectable difference.6

Latitude (Colombia) November–April May–October Source
4–5° N 60% of flowers 40% Unigarro et al. 20236
8–9° N 90% 10% Unigarro et al. 20236
Above 10° N All flowering None Unigarro et al. 20236

The 2023 study gives these shares in its introduction, from earlier literature.6 Outside the equatorial zone, a 2007 review describes flowering in south-central Brazil in two to four or more synchronised blossom periods.2 Farther from the equator, more than 10° N or S, the 2023 study cites one to three flowering events a year.6 The Semperflorens mutation flowers continuously through the year in Brazil.1

Abnormal flowers#

“Star flowers” are flowers that open prematurely, with small, stiff petals in a starlet shape.3 They form under two opposite stresses. High temperatures with intense water deficit at the start of anthesis dehydrate the pollen tube and cause floral atrophy and abortion.3,1 Prolonged water excess has the same result by restarting dormant buds before they are conditioned by a dry period.1 In high-rainfall areas and La Niña years, star flowers can make up 12.9–17.8% of recorded flowers.1 A 2017 textbook chapter lists high temperature with high rainfall as a cause of flower atrophy and poor fruit set.8

Pollination biology#

After anthesis, pollen stays viable for 24–36 h and the stigma stays receptive for 48 h.1 Initial fruit set under natural pollination was 75.2% at 35 days after flowering; final fruit set has ranged from 50% to 57%.1 Insect-mediated cross-pollination gave 9.4–26.7% of fruit set in Colombia, close to the 10–30% reported in the wider literature.1

In a 2003 pollination study in Sulawesi, Indonesia, self-pollinated arabica flowers set 60% fruit; open-pollinated flowers set 75.2%.10 In canephora, self-pollinated flowers and unmanipulated controls set 10% or less.10 Wind pollination, including self-pollination, gave fruit set 16 percentage points lower than open pollination in canephora and 12.3 points lower in arabica.10

Parameter Value Source
Dormancy of mature buds 30–120 days Unigarro et al. 20251
Rain to anthesis, C. arabica 10–14 days; 9–12 days after re-irrigation (Hawaii) Rendón-Sáenz et al.4; Crisosto et al.5
Rain to anthesis, C. canephora 8–11 days Rendón-Sáenz et al.4
Rain to anthesis, C. liberica 7–10 days Rendón-Sáenz et al.4
Flowers per inflorescence (C. arabica) 4–5 Unigarro et al. 20251
Flower open 2–3 days Unigarro et al. 20251
Pollen viability 24–36 h Unigarro et al. 20251
Stigma receptivity 48 h Unigarro et al. 20251
Cross-pollinated share of fruit set (C. arabica, Colombia) 9.4–26.7% Unigarro et al. 20251
Fruit set, self-pollinated (C. arabica / C. canephora) 60% / 10% or less Klein et al. 200310
Star flowers, high-rainfall areas 12.9–17.8% of flowers Unigarro et al. 20251

Self-incompatibility in canephora#

C. canephora has gametophytic self-incompatibility, controlled by a single multiallelic gene, the S gene.11 In this system, the pollen grain must not share its allele with the receiving plant. Pollen tubes that do stop growing, through the action of ribonucleases that degrade their ribosomal RNA.11 Plants with the same S genotype form one compatibility group and cannot pollinate each other.11

A 2018 Brazilian study placed 12 selected clones in three compatibility groups with a balanced diallel of controlled crosses. Compatible crosses set fruit at a mean rate of 44%.11 The authors note that planting incompatible clones together can compromise yield; knowing each clone’s S genotype lets farms plant compatible clones side by side.11 The 2003 study notes that self-sterility in canephora could also reflect inbreeding depression after self-pollination.10

Pollination services#

In Sulawesi, Indonesia, fruit set across 24 agroforestry plots was predicted by the number of flower-visiting bee species, not by bee abundance.12 Single visits by solitary bees gave higher fruit set than single visits by social bees (87.3% against 74.7%), but social bees visited more flowers.12

In Costa Rica, a study on one large farm of Caturra arabica measured pollination along distance gradients from forest fragments.13 Forest-based pollinators raised yields by 20% within 1 km of forest and cut the share of peaberries by 27%.13 For 2000–2003, the authors valued the service from two forest fragments (46 and 111 ha) at USD 60,000 per year for the one farm (the results section gives USD 62,000).13 Earlier work cited in the same paper found that bee visits raise arabica yields by 15–50% over flowers with bees excluded.13 Peaberries are covered in Peaberries and bean shapes.

Caffeine in nectar may help keep pollinators faithful. In a 2013 study, caffeine, which occurs in Coffea and Citrus nectar, strengthened the responses of honeybee neurons involved in olfactory learning and memory.14 The concentrations did not exceed the bees’ bitter-taste threshold.14 The authors interpret this as a way for plants to secure pollinator fidelity; the chemistry is covered in Caffeine.14

Open questions#

  • Rain threshold. No source opened for this entry measured, at several sites with one method, the rainfall needed to open dormant buds; published figures range from 5 mm to more than 14 mm.1,4
  • Canephora flowering physiology. The reviews opened for this entry deal mainly with C. arabica; for canephora they give rain-to-anthesis intervals but little on induction and dormancy.4
  • Bee pollination in arabica. Field studies agree that bees raise arabica fruit set, but the share of fruit set due to cross-pollination ranges from about 10% to 30% between studies and sites.1

Key studies#

Crisosto et al., 1992 – Water deficit and flower opening. Dried arabica plants to set water potentials and re-watered them in greenhouses and in the field in Hawaii. Flowers opened only after a deficit below −0.8 MPa, or a milder one held for two weeks, and only from buds at one developmental stage; a root signal appeared to drive the response.

Tree Physiology 10(2), 127–139 · doi:10.1093/treephys/10.2.127

Klein et al., 2003 – Bee pollination in arabica and canephora. Compared self-, cross-, open and wind pollination in both species. Canephora set almost no fruit from its own pollen; arabica did, but open and cross-pollination gave the highest fruit set in both.

American Journal of Botany 90(1), 153–157 · doi:10.3732/ajb.90.1.153

Ricketts et al., 2004 – The value of forest pollinators. Measured arabica pollination on a Costa Rican farm at different distances from forest. Near forest, yields were 20% higher and peaberries 27% fewer, which the authors valued at about USD 60,000 a year for the farm.

PNAS 101(34), 12579–12582 · doi:10.1073/pnas.0405147101

Unigarro et al., 2025 – Arabica phenology from flower to fruit. A review of flowering and fruiting in arabica on the BBCH scale, with bud dormancy, rain triggers, thermal time, flower structure, pollination and fruit set, drawing heavily on Colombian research.

Plants 14(21), 3396 · doi:10.3390/plants14213396

Rendón-Sáenz et al., 2025 – Triggers for coffee flowering. A systematic review of endogenous and environmental flowering triggers. It collects rain thresholds, rain-to-anthesis intervals for three species, temperature limits and photoperiod effects across producing countries.

Frontiers in Sustainable Food Systems 9, 1711697 · doi:10.3389/fsufs.2025.1711697

Contested
  • Rain needed to open dormant buds: a 2025 review gives 5–10 mm to restart bud growth, with more than 14 mm possibly needed for a full flowering1; a 2025 systematic review gives more than 10 mm in a single day after a dry period.4 The studies behind these figures used different sites and definitions.
  • Water deficit that prepares buds: a Hawaii experiment found that re-irrigation opened flowers only if predawn leaf water potential had fallen below −0.8 MPa, or stayed at about −0.3 to −0.5 MPa for two weeks5; a 2007 review cites a threshold of about −1.2 MPa2; a 2025 review gives soil water potentials below −1.4 MPa.1 The three figures measure different things (leaf or soil) under different conditions.
  • Upper temperature for flower initiation: a 2025 review states that initiation does not occur above 28 °C or with more than 13 h of daylight1; a 2025 systematic review cites day temperatures above 30 °C with nights above 17 °C as inhibiting it.4
Share or cite this entry

Change log

  • – Colombian flowering shares now given for November–April and May–October instead of the first and second half of the year. Reason: Unigarro et al. 2023 define the two flowering periods as November–April (FP I) and May–October (FP II).
  • – The Nicaragua intervals (8–10 and 10–13 days) now describe large and smaller flowerings after rain, not heavy and light rain; text and chart labels corrected. Reason: Table 3 of Rendón-Sáenz et al. 2025 attributes the 10–13 days to a less intense flowering event.
  • – The −1.2 MPa figure is now described as a threshold for triggering flower opening cited from a 1976 study, not as the point to resume watering. Reason: DaMatta et al. 2007 give it as a critical value for triggering flower opening.

Sources

  1. AUnigarro, C. A.; Cayón Salinas, D. G.; León-Burgos, A. F.; Flórez-Ramos, C. P. (2025). Flowering and Fruiting of Coffea arabica L.: A Comprehensive Perspective from Phenology. Plants 14(21), 3396. doi:10.3390/plants14213396 · other copies
  2. ADaMatta, F. M.; Ronchi, C. P.; Maestri, M.; Barros, R. S. (2007). Ecophysiology of coffee growth and production. Brazilian Journal of Plant Physiology 19(4), 485–510. doi:10.1590/S1677-04202007000400014 · other copies
  3. ALópez, M. E.; Santos, I. S.; de Oliveira, R. R.; Lima, A. A.; Cardon, C. H.; Chalfun-Junior, A. (2021). An overview of the endogenous and environmental factors related to the Coffea arabica flowering process. Beverage Plant Research 1, 13. doi:10.48130/BPR-2021-0013 · other copies
  4. ARendón-Sáenz, J. R.; Zamorano-Montañez, C.; García-López, J. C.; Ceballos-Aguirre, N.; Castiblanco-Carranza, M. A.; Vargas-López, J. (2025). Unveiling triggers for flowering in coffee plants: a systematic review of endogenous and environmental factors. Frontiers in Sustainable Food Systems 9, 1711697. doi:10.3389/fsufs.2025.1711697 · other copies
  5. ACrisosto, C. H.; Grantz, D. A.; Meinzer, F. C. (1992). Effects of water deficit on flower opening in coffee (Coffea arabica L.). Tree Physiology 10(2), 127–139. doi:10.1093/treephys/10.2.127 · other copies
  6. AUnigarro, C. A.; Imbachi, L. C.; Darghan, A. E.; Flórez-Ramos, C. P. (2023). Quantification and Qualification of Floral Patterns of Coffea arabica L. in Colombia. Plants 12(18), 3332. doi:10.3390/plants12183332 · other copies
  7. AWormer, T. M.; Gituanja, J. (1970). Floral Initiation and Flowering of Coffea arabica L. in Kenya. Experimental Agriculture 6(2), 157–170. doi:10.1017/S0014479700000211 · other copies
  8. BLambot, C.; Herrera, J. C.; Bertrand, B.; Sadeghian, S.; Benavides, P.; Gaitán, A. (2017). Cultivating Coffee Quality – Terroir and Agro-Ecosystem. In: Folmer, B. (ed.), The Craft and Science of Coffee, Academic Press, ch. 2, pp. 17–49. doi:10.1016/B978-0-12-803520-7.00002-5 · other copies
  9. AScalabrin, S.; Toniutti, L.; Di Gaspero, G.; et al.; Bertrand, B. (2020). A single polyploidization event at the origin of the tetraploid genome of Coffea arabica is responsible for the extremely low genetic variation in wild and cultivated germplasm. Scientific Reports 10, 4642. doi:10.1038/s41598-020-61216-7 · other copies
  10. AKlein, A.-M.; Steffan-Dewenter, I.; Tscharntke, T. (2003). Bee pollination and fruit set of Coffea arabica and C. canephora (Rubiaceae). American Journal of Botany 90(1), 153–157. doi:10.3732/ajb.90.1.153 · other copies
  11. AMoraes, M. S.; Teixeira, A. L.; Ramalho, A. R.; Espíndula, M. C.; Ferrão, M. A. G.; Rocha, R. B. (2018). Characterization of gametophytic self-incompatibility of superior clones of Coffea canephora. Genetics and Molecular Research 17(1), gmr16039876. doi:10.4238/gmr16039876 · other copies
  12. AKlein, A.-M.; Steffan-Dewenter, I.; Tscharntke, T. (2003). Fruit set of highland coffee increases with the diversity of pollinating bees. Proceedings of the Royal Society B 270(1518), 955–961. doi:10.1098/rspb.2002.2306 · other copies
  13. ARicketts, T. H.; Daily, G. C.; Ehrlich, P. R.; Michener, C. D. (2004). Economic value of tropical forest to coffee production. Proceedings of the National Academy of Sciences 101(34), 12579–12582. doi:10.1073/pnas.0405147101 · other copies
  14. AWright, G. A.; Baker, D. D.; Palmer, M. J.; Stabler, D.; Mustard, J. A.; Power, E. F.; Borland, A. M.; Stevenson, P. C. (2013). Caffeine in floral nectar enhances a pollinator's memory of reward. Science 339(6124), 1202–1204. doi:10.1126/science.1228806 · other copies

Tier A · authoritative · B · trade and expert · C · general. Supporting quotes for every source are in the data version of this entry. If a publisher’s site refuses access, try the doi link or “other copies”.

Suggest a change

On Flowering and pollination. Tell us what you would change and why. A source that supports it helps most.