Nutrition and fertilisation

Removal by the harvest, measured#

Whole fruit, Colombia. Cenicafé expresses removal per 1,000 kg of green coffee (café almendra), equal to 1,250 kg of dry parchment coffee.¹ The whole fruit behind that tonne held 30.9 kg N, 2.3 kg P, 36.9 kg K, 4.3 kg Ca, 2.3 kg Mg and 1.2 kg S.¹ Micronutrients were 107 g Fe, 61 g Mn, 50 g B, 33 g Cu and 18 g Zn.¹ Split by fraction, the green coffee held 16.79 kg N and 12.40 kg K, and the pulp 11.50 kg N and 22.31 kg K.¹ Parchment and mucilage held the rest.¹

Green coffee alone, Brazil. A 1963 study in São Paulo compared Mundo Novo, Caturra and Bourbon Amarelo on three soils.² The varieties did not differ.² The beans held 1.71% N, 0.10% P, 1.53% K, 0.27% Ca, 0.15% Mg and 0.12% S in the dry matter.² That is 17.1 kg N and 15.3 kg K per tonne of bean dry matter (our calculation), of the same order as the Colombian bean fraction (16.79 kg N and 12.40 kg K per 1,000 kg green coffee), although the bases differ.¹ By the authors’ estimate, the clean coffee carried about two-thirds of the fruit’s nitrogen, half of its phosphorus and potassium, and one-third of its calcium, magnesium and sulphur.² The rest stays in pulp and husk, which the authors proposed returning to the field.²

Robusta, Brazil. In a 2014/2015 trial with C. canephora (Conilon BRS Ouro Preto) in Rondônia, western Amazon, fertiliser doses from 0 to 3,000 kg per hectare and year of a 20-00-20 formulation did not change nutrient concentrations in the fruit.³ Higher doses raised dry mass, and so total export.³ Across all doses the beans exported N > K > Mg > P > Ca > S; the husk exported K > N > Ca > Mg > S > P.³ From the trial means (4,455 kg bean dry matter and 121.73 kg N per hectare), the beans held about 27.3 kg N and 20.5 kg K per tonne of dry matter (our calculation).³ According to the 2017 textbook chapter, removal per tonne of marketable coffee is of the same order for robusta and arabica.⁴

Pulp as a nutrient source. Producing 1,250 kg of dry parchment coffee leaves about 2,700 kg of fresh pulp.¹ Cenicafé puts its content at about 10.2 kg N, 1.4 kg P₂O₅, 23.8 kg K₂O, 2.2 kg CaO, 0.8 kg MgO and 0.4 kg SO₄.¹ It recommends 6–12 kg of decomposed pulp per tree and year as an organic fertiliser.¹

Nutrients and deficiency signs#

According to Cenicafé, nitrogen is the nutrient that most limits production in Colombia.¹ Leaving nitrogen out of fertilisation lowered yield by up to 80%.¹ Potassium ranks second; on soils very low in potassium (below 0.2 cmol_c/kg), yield fell by nearly 30%.¹ Responses to phosphorus were occasional and small, mainly below 10 mg/kg soil phosphorus.¹ Magnesium deficiency is common in the leaves of productive branches; sulphur deficiency is rarely seen.¹

The chapter links nitrogen to chlorophyll formation.¹ It does not describe the roles of the other nutrients, and no source consulted for this entry measured them in coffee. Deficiency symptoms, as Cenicafé lists them:

Nitrogen and sulphur both cause even yellowing; the age of the affected leaves separates them.¹ According to the 2017 textbook chapter, micronutrient deficiencies are identified by such symptoms or by leaf analysis.⁴

Leaf and soil analysis#

Leaf sufficiency ranges. Cenicafé gives normal leaf concentrations for Colombia: N 2.30–2.80%, P 0.10–0.18%, K 1.50–2.00%, Ca 0.50–1.30% and Mg 0.30–0.40%.¹ For micronutrients it gives Mn 150–220 mg/kg and Fe 90–140 mg/kg; its boron range of 40–60 carries no unit in the table.¹ A 2003 Brazilian study compiled six published ranges beside its own.⁵ Their limits for leaf nitrogen run from 2.30% to 3.50%, and for potassium from 1.80% to 2.60%.⁵

Regional ranges in Minas Gerais. The 2003 study sampled 312 plots on 168 farms in four regions, in a high-yield and a low-yield year.⁵ It took farms yielding more than 1,800 kg of green coffee per hectare, averaged over the two years, as the reference.⁵ The critical ranges did not differ between years or regions.⁵ Micronutrient problems were the most frequent finding.⁵

Soil classes. For producing coffee, Cenicafé classes soil phosphorus as low below 10 mg/kg and high above 20 mg/kg.¹ Potassium is low below 0.2 cmol_c/kg, calcium below 1.5 and magnesium below 0.6 cmol_c/kg.¹ Cenicafé considers a soil pH of 5.0–5.5 adequate, and toxicity becomes more likely when exchangeable aluminium exceeds 1 cmol_c/kg.¹ Liming is covered in soil acidity and liming.

Programmes by age and yield#

Colombia (Cenicafé). For young plants, nitrogen ranges from 5 to 16 g per plant, depending on age (2–18 months) and soil organic matter.¹ Phosphorus (4–6 g P₂O₅), potassium (5–10 g K₂O) and magnesium (2–3 g MgO) per plant are added only where soil tests fall below set thresholds.¹ For producing coffee with a soil analysis, the recommendation is 120–300 kg N, 0–300 kg potassium, 0–60 kg each of magnesium and phosphorus, and up to 50 kg sulphur per hectare and year.¹ Without an analysis, Cenicafé suggests 300 kg N, 260 kg K₂O and 50 kg each of P₂O₅, MgO and S for more than 7,500 plants per hectare in full sun.¹ Doses fall with lower planting density and with shade.¹

Brazil, dense plantings (trial optima). At Martins Soares, Minas Gerais (763 m), Catuaí Vermelho 44 at 9,523 plants per hectare received 0–1,200 kg per hectare and year of N and K₂O in a 1:1 ratio.⁶ Over eight seasons, the dose of maximum technical efficiency was 881.7 kg N + K₂O, for a maximum mean yield of 61.9 bags per hectare.⁶ The paper does not state the bag weight; with the usual 60 kg bag of green coffee, that is about 3,714 kg per hectare (our calculation). In a second trial with Catucaí Amarelo 6/30 at 6,666 plants per hectare, yield did not respond in the first two seasons.⁶ In the four later seasons, the doses for maximum yield were 389–560 kg N and 21–373 kg K₂O per hectare, varying by season (the abstract rounds this to 424–560 kg N and 21–338 kg K₂O).⁶ Nitrogen reduced the swing between heavy and light crops (biennial bearing).⁶ These are agronomic optima of two trials, not general recommendations.

Other producing countries. No technical recommendation of an African or Asian coffee research institute was consulted for this entry. Doses here are therefore from Colombia and Brazil only.

Timing, splitting and foliar feeding#

With the rains. Cenicafé ties applications to soil moisture.¹ In dry conditions or El Niño years it prefers two applications about six months apart; at wet sites or in La Niña years, three, about four months apart.¹ Where rain falls in two seasons, fertiliser goes on at the start of the wetter periods; with one rainy season, at its onset.¹ Fertilisers applied at low volume, such as diammonium phosphate or magnesium oxide, can go on in one dose a year.¹ Urea is very soluble in water and at greater risk of loss.¹

Uptake over the year. In a Brazilian ¹⁵N study, nitrogen was applied as ammonium sulphate in four splits.⁸ By February of the first year, the aerial parts held 71% of the fertiliser nitrogen applied.⁸ The share fell to 34% at harvest and 19% at the next flowering, as leaves fell and fruit was removed.⁸ The authors found the four-split programme adequate for the crop’s needs.⁸ Fruit takes up most of its nutrients two to six months after flowering, according to Cenicafé.¹ How the fruit grows in that time is covered in fruit development.

Foliar feeding. According to the 2017 textbook chapter, foliar sprays are often used to correct deficiencies of zinc, boron, iron and manganese, identified by symptoms or leaf analysis.⁴ Cenicafé states that a beneficial effect of foliar fertilisers has not yet been shown, and leaves them out of its recommendations.¹ No study consulted for this entry measured yield or uptake from foliar sprays on coffee.

Nitrogen losses and the environment#

A ¹⁵N balance in Brazil. In Piracicaba, São Paulo, coffee received 280 and 350 kg N per hectare in two seasons as ¹⁵N-labelled ammonium sulphate.⁷ After two years, 19.1% of the fertiliser nitrogen was in the aerial parts, 9.4% in roots, 23.8% in litter and 26.3% in fruit.⁷ The soil to 1 m held 12.6%.⁷ Leaching lost 2.0% and volatilisation 0.9% a year, which the authors called very small; 6.2% was not accounted for.⁷

Leaching in Costa Rica. Shaded and unshaded plantations in the Central Valley both received 300 kg N per hectare and year.¹⁰ Measured with ceramic-cup lysimeters, nitrate leaching was 24 kg nitrate nitrogen per hectare and year without shade and 9 kg with shade.¹⁰ Denitrification was higher in shaded soils, and the authors concluded that in both plantation types the potential for loss by leaching was small compared with that for gaseous loss.¹⁰

Nitrous oxide. On Andosols in Costa Rica, a coffee monoculture and coffee shaded by Inga densiflora both received 250 kg N per hectare and year.⁹ Over one year, the monoculture emitted 4.3 ± 0.1 kg and the shaded system 5.8 ± 0.3 kg N₂O nitrogen per hectare.⁹ Of the yearly N₂O, 84% was emitted in the periods after fertilisation.⁹ Litterfall added 153 kg N in the monoculture and 246 kg N under Inga.⁹ The authors concluded that nitrogen-fixing shade trees can increase N₂O emissions.⁹ Shade trees as a nitrogen source are discussed in shade and agroforestry.

Footprint. In long-term trials in Costa Rica and Nicaragua, organic and inorganic nitrogen inputs were the main contributor to greenhouse gas emissions in every management system.¹¹ No study consulted for this entry measured nitrate in drinking water near coffee farms.

Nutrition, disease and cup quality#

Rust. In a controlled-environment study, high light combined with low nitrogen fertilisation changed the course of coffee leaf rust and led to heavy sporulation.¹² The authors proposed vigorous varieties, shade and appropriate nitrogen fertilisation as part of an agroecological approach to the disease.¹² The disease itself is covered in coffee leaf rust.

Brown eye spot and rust under organic fertilisers. A two-year Brazilian field trial (2003–2005) compared organic nutrient sources.¹³ Coffee fruit peel applied alone gave the greatest disease progress: 62% above the control for brown eye spot (Cercospora coffeicola) and 38% above it for rust.¹³ These plants had more potassium and less calcium in their leaves, lost more leaves and yielded less.¹³ Mixing the peel with castor bean presscake or swine waste cut brown eye spot by 38% compared with peel alone and raised leaf calcium.¹³ The authors concluded that better nutritional balance increased resistance to the diseases.¹³

Cup quality. The 2017 textbook chapter summarises older studies that disagree.⁴ Some found no effect of fertilisation on cup quality; others found that high nitrogen or potassium doses, or some fertiliser forms, lowered it.⁴ The chapter concludes that both negative and positive effects occur and that more research is needed.⁴ No study consulted for this entry tested fertilisation and cup quality directly.

Reference values#

Open questions#

• Recommendations beyond Latin America. This entry rests on Colombian and Brazilian work. No technical recommendation from an African or Asian coffee research institute was consulted for it.
• Foliar feeding. Whether foliar sprays raise yield is not established in the sources consulted here; they disagree on its place in a programme.^1,4
• Nutrient roles in coffee. No source consulted for this entry measured the physiological roles of single nutrients, other than nitrogen, in coffee.
• Fertilisation and cup quality. Older studies disagree, and the 2017 textbook chapter calls for more research.⁴
• Gaseous versus leaching losses. The two field studies consulted here weigh them differently, on different soils and with different fertilisers.^7,10

Key studies#