{ "id": "soil-acidity-and-liming", "url": "https://coffearium.org/entries/soil-acidity-and-liming/", "license": "CC-BY-4.0", "title": "Soil acidity and liming", "summary": "Why coffee soils turn acid, how aluminium and manganese harm the roots, and how lime and gypsum are used to correct the topsoil and the subsoil, with the rates that research institutes and field trials report.", "domain": "farm-management", "kind": "Practice", "stage": "farm", "topics": [ "science" ], "topicSections": {}, "lang": "en", "status": "verified", "reviewer": "Coffearium verification pass", "lastVerified": "2026-10-23T00:00:00.000Z", "updated": "2026-10-23T00:00:00.000Z", "added": "2026-10-23T00:00:00.000Z", "brief": [ "The Colombian coffee research centre Cenicafé gives pH 5.0–5.5 as the suitable soil range for coffee; below pH 5.0 the plant grows less.2", "In nutrient solution, aluminium reduced shoot, root and fruit weight of coffee even at the lowest concentration tested.4", "In field trials at two sites in Paraná, Brazil (1975–82), 2.5 t/ha of dolomitic lime gave the best yields; 5 and 10 t/ha gave less.6", "Lime acts mainly where it lies; its effects stayed in the top 30 cm in a Brazilian coffee trial and the top 10–20 cm in a grain-crop trial.6,7", "Gypsum carries calcium deeper; in a Brazilian grain-crop trial it lowered the aluminium saturation of the subsoil, with little change in pH.2,7" ], "quickFacts": [ { "label": "Suitable soil pH (Colombia)", "value": "5.0–5.5 (Cenicafé)" }, { "label": "Brazilian lime method", "value": "Raise base saturation to the target set in each state's manual" }, { "label": "Best rate, Paraná trial 1975–82", "value": "2.5 t/ha dolomitic lime; 5 and 10 t/ha yielded less" }, { "label": "Depth reached by surface lime", "value": "0–30 cm (coffee trial); 10–20 cm after 3–6 years (no-till trial)" }, { "label": "Liming, Indian robusta estates", "value": "20% more harvested berries per infructescence on estates that limed" }, { "label": "Leaf aluminium with visible injury", "value": "Above 300 µg/g (nutrient-solution study)" } ], "aka": [ "Liming", "Soil acidity", "Liming coffee soils" ], "linkTerms": [ "liming", "lime requirement", "agricultural lime", "dolomitic lime", "dolomitic limestone", "calcitic lime", "soil acidity", "acid soils", "aluminium toxicity", "aluminum toxicity", "agricultural gypsum", "gypsum application", "base saturation" ], "related": [ "soils-for-coffee", "nutrition-and-fertilisation", "site-selection-and-climate", "shade-and-agroforestry", "coffea-arabica", "coffea-canephora", "irrigation", "coffee-yields" ], "contested": [ "High gypsum rates: in a Minas Gerais trial, 14 t/ha gave the highest four-year yield; 56 t/ha yielded less than 14 t/ha but more than no gypsum, and the authors call it excessive. The near doubling of yield in the first year was not maintained.9 The paper says farmers apply gypsum above the recommended rate; the trial alone does not settle an optimum.9" ], "corrections": [ { "date": "2026-10-23T00:00:00.000Z", "change": "Melo & Sartori 2013, Costa et al. 2020 and Lange et al. 2025 set to Tier B (peer review not stated); the 60%, 70% and 50% base saturation targets and the 90-day depth result removed.", "reason": "A single Tier B source cannot carry a number." }, { "date": "2026-10-23T00:00:00.000Z", "change": "The 56 t/ha gypsum rate now reported as yielding less than 14 t/ha but more than no gypsum, instead of \"reduced yield\".", "reason": "Cogo et al. 2020, Table 4 (255 sacks/ha at 56 t/ha against 238 without gypsum)." }, { "date": "2026-10-23T00:00:00.000Z", "change": "The Chaves et al. 1984 lime trials described as two sites between 1975 and 1982, effects mainly in the top 30 cm, instead of one eight-year trial.", "reason": "Full text of Chaves et al. 1984 (Londrina 1974–79, Florestópolis 1979–82)." }, { "date": "2026-10-23T00:00:00.000Z", "change": "Grain-crop evidence (Pauletti et al. 2014) labelled as such in the brief, figure legend and caption; the 0.29 pH rise reattributed to stopping nitrogen in a separate 41-lot study.", "reason": "Pauletti et al. 2014 is not a coffee trial; Cenicafé Avance Técnico 466." } ], "sources": [ { "id": "s1", "title": "Agroforestry coffee production increased by native shade trees, irrigation, and liming", "authors": "Boreux, V.; Vaast, P.; Madappa, L. P.; Cheppudira, K. G.; Garcia, C.; Ghazoul, J.", "year": 2016, "publisher": "Agronomy for Sustainable Development 36(3), 42", "url": "https://link.springer.com/article/10.1007/s13593-016-0377-7", "doi": "10.1007/s13593-016-0377-7", "tier": "A", "quote": "Irrigation and liming increased berry production respectively by 16 and 20 %.", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s2", "title": "La acidez del suelo, una limitante común para la producción de café", "authors": "Sadeghian Khalajabadi, S.", "year": 2016, "publisher": "Cenicafé, Avance Técnico 466", "url": "https://www.cenicafe.org/es/publications/avt0466.pdf", "tier": "A", "tierNote": "Research publication of the national coffee research centre of Colombia.", "quote": "Para café el rango adecuado de pH se encuentra entre 5,0 y 5,5", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s3", "title": "Toxicity of Aluminum to Coffee Seedlings Grown in Nutrient Solution", "authors": "Pavan, M. A.; Bingham, F. T.", "year": 1982, "publisher": "Soil Science Society of America Journal 46(5), 993–997", "url": "https://agris.fao.org/search/en/records/65dfc4d40f3e94b9e5dc27a8", "doi": "10.2136/sssaj1982.03615995004600050021x", "tier": "A", "quote": "was associated with a 10% reduction of coffee seedling growth", "quotes": [], "abstractOnly": true, "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s4", "title": "Aluminum toxicity in coffee trees cultivated in nutrient solution", "authors": "Pavan, M. A.; Bingham, F. T.", "year": 1982, "publisher": "Pesquisa Agropecuária Brasileira 17(9), 1293–1302", "url": "https://apct.sede.embrapa.br/pab/article/view/16275", "doi": "10.1590/S1678-3921.pab1982.v17.16275", "tier": "A", "quote": "There was significant reduction in all growth parameters even at lowest Al treatment.", "quotes": [], "abstractOnly": true, "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s5", "title": "Toxicity of metals in plants. I. Characterization of manganese toxicity in coffee", "authors": "Pavan, M. A.; Bingham, F. T.", "year": 1981, "publisher": "Pesquisa Agropecuária Brasileira 16(6), 815–821", "url": "https://apct.sede.embrapa.br/pab/article/view/17119", "doi": "10.1590/S1678-3921.pab1981.v16.17119", "tier": "A", "quote": "Manganese toxicity symptoms were marginal chlorosis and few necrotic spots on younger leaves.", "quotes": [], "abstractOnly": true, "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s6", "title": "Response of coffee to lime (Respostas do cafeeiro à calagem)", "authors": "Chaves, J. C. D.; Pavan, M. A.; Igue, K.", "year": 1984, "publisher": "Pesquisa Agropecuária Brasileira 19(5), 573–582", "url": "https://apct.sede.embrapa.br/pab/article/view/15714", "doi": "10.1590/S1678-3921.pab1984.v19.15714", "tier": "A", "quote": "The best coffee yields were obtained with the lowest lime rate (2.5 tons/ha).", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s7", "title": "Efeitos em longo prazo da aplicação de gesso e calcário no sistema de plantio direto (Long-term effects of the application of gypsum and lime in a no-till system)", "authors": "Pauletti, V.; de Pierri, L.; Ranzan, T.; Barth, G.; Motta, A. C. V.", "year": 2014, "publisher": "Revista Brasileira de Ciência do Solo 38(2), 495–505", "url": "https://www.redalyc.org/pdf/1802/180231134014.pdf", "tier": "A", "tierNote": "Field trial with grain crops, not coffee; cited for how lime and gypsum move in a Brazilian Latosol.", "quote": "Liming amended pH, increased the concentration of Ca and Mg and decreased Al in the surface layers of the soil.", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s8", "title": "Avaliação da escória de siderurgia e do calcário como corretivos para cultura cafeeira", "authors": "Melo, B. M. R.; Sartori, R. H.", "year": 2013, "publisher": "Revista Agrogeoambiental 5(1), 11–18", "url": "https://agrogeoambiental.ifsuldeminas.edu.br/index.php/Agrogeoambiental/article/download/432/450/890", "tier": "B", "tierNote": "Journal of the Federal Institute of Southern Minas Gerais; the article lists an evaluation and copy-editing period but peer review is not stated, so Tier B. Carries no number here; cited only for statements about its own trial.", "quote": "A dose aplicada foi em função da elevação da saturação de bases do solo para 60% (V%)", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s9", "title": "High rates of agricultural gypsum affect the arbuscular mycorrhiza fungal community and coffee yield", "authors": "Cogo, F. D.; Saggin Júnior, O. J.; Guimarães, P. T. G.; Siqueira, J. O.; Carneiro, M. A. C.", "year": 2020, "publisher": "Bragantia 79(4), 487–497", "url": "https://www.alice.cnptia.embrapa.br/alice/bitstream/doc/1126810/1/High-rates-of-agricultural-gypsum.pdf", "doi": "10.1590/1678-4499.20200014", "tier": "A", "quote": "Coffee yield was reduced with the highest gypsum rate (56 t·ha-1)", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s10", "title": "Uso do calcário e do gesso agrícola em duas épocas de implantação do Coffea arabica L.", "authors": "Costa, B. P.; Duarte Júnior, J. B.; Rego, C. A. R. M.; Costa, A. C. T.; Lana, M. C.", "year": 2020, "publisher": "Revista Brasileira de Agropecuária Sustentável 10(1), 241–247", "url": "https://periodicos.ufv.br/rbas/article/download/9940/6121/50691", "tier": "B", "tierNote": "Received and accepted dates are given but peer review is not stated in the article, so Tier B. Carries no number here.", "quote": "O calcário e o gesso agrícola não influenciaram nas características agronômicas nas duas épocas de avaliação", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s11", "title": "Curvas de neutralização da acidez de solos na transição Cerrado-Amazônia", "authors": "Lange, A.; Contesini, M. E.; Moraes, W. C. P.", "year": 2025, "publisher": "Nativa 13(2), 290–300", "url": "https://periodicoscientificos.ufmt.br/ojs/index.php/nativa/article/download/19143/14729/89918", "doi": "10.31413/nat.v13i2.19143", "tier": "B", "tierNote": "Soils under grain crops, not coffee; submission and acceptance dates are given but peer review is not stated in the article, so Tier B. Cited with the Embrapa handbook for the base saturation formula; carries no number.", "quote": "NC= (V2-V1)*T/PRNT", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s12", "title": "Cambios químicos ocasionados por encalamiento y uso de lombrinaza en la etapa de almácigo del café", "authors": "Díaz-Marín, C.; Sadeghian-Khalajabadi, S.; Morales-Londoño, C. S.", "year": 2008, "publisher": "Cenicafé 59(4), 295–309", "url": "https://www.cenicafe.org/es/publications/arc059%2804%29295-309.pdf", "tier": "A", "quote": "Liming is the most recommended practice for soil acidity management.", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" }, { "id": "s13", "title": "Calagem e adubação. In: Gonzaga, A. C. de O. (ed.), Feijão: o produtor pergunta, a Embrapa responde, 2nd ed., pp. 59–76", "authors": "Fageria, N. K.; Carvalho, M. da C. S.; Oliveira, I. P. de", "year": 2014, "publisher": "Embrapa, Brasília (Coleção 500 Perguntas, 500 Respostas)", "url": "https://www.infoteca.cnptia.embrapa.br/infoteca/bitstream/doc/1015113/1/p59.pdf", "tier": "B", "tierNote": "Extension handbook for common beans; cited only for how the Brazilian lime-requirement methods are built, never for a number.", "quote": "Existem três métodos para determinar a necessidade de calcário", "quotes": [], "accessed": "2026-10-23T00:00:00.000Z" } ], "markdown": "## What soil acidity is\n\nSoil pH measures how acid a soil is. A lower number means a more acid soil. For coffee, the Colombian coffee research centre (Cenicafé) gives a suitable range of pH 5.0–5.5.2 Below pH 5.0 the plant grows less, and the same happens above 5.5, according to Cenicafé.2\n\nIn acid soils, aluminium ions (Al³⁺) take the place of calcium, magnesium and potassium on the soil particles.2 These three nutrients are called bases. At high concentrations, the exchangeable aluminium is toxic to plants.2 Manganese can also reach toxic levels.2 Above pH 5.5 the aluminium is neutralised and stops limiting coffee growth.2\n\nAcid soils are common in coffee land. In the soil-test records of Antioquia, Colombia, 79% of 69,718 samples had a pH of 5 or less.2 How texture, depth and organic matter shape a coffee soil is the subject of Soils for coffee.\n\n## Why coffee soils turn acid\n\nRain is the first cause. Water moving through the soil washes the bases below the reach of the roots.2 Acid cations, mainly aluminium, take their place.2 Cenicafé also names the decomposition of organic matter and the oxidation of sulphur.2\n\nNitrogen fertiliser is the second cause. When soil microbes turn ammonium into nitrate (nitrification), acid is released.2 Cenicafé calls nitrification one of the most important processes in soil acidification.2 Per unit of nitrogen, ammonium sulphate and monoammonium phosphate acidify twice as much as urea or ammonium nitrate.2\n\nIn a four-year Colombian trial, 300 kg of nitrogen per hectare and year lowered soil pH by 0.27 units.2 How much nitrogen coffee needs is covered in Nutrition and fertilisation.\n\n## What acidity does to coffee roots\n\nAluminium harms the roots first. In a two-year greenhouse study, coffee plants grew in nutrient solutions with six aluminium levels.4 Shoot, root and fruit weight fell progressively as aluminium rose.4 Growth was lower even at the lowest level tested.4\n\nThe roots of these plants became thicker and shorter, with fewer side roots.4 Young leaves turned small, pale and curled at the edges.4 The plants took up less calcium, magnesium and phosphorus.4\n\nToo much manganese gives a different picture. In a sand-culture study, its main symptoms were pale leaf edges and a few dead spots on young leaves.5\n\n## Lime and gypsum: two different jobs\n\nLime is ground rock made mainly of calcium carbonate, magnesium carbonate or both.2 Dolomitic lime contains both: the one used in a Brazilian trial had 20.1% magnesium oxide.2,6 Liming raises pH and exchangeable calcium and magnesium.6 Cenicafé lists among its benefits that it supplies calcium and magnesium and removes aluminium and manganese toxicity.2 A 2008 Cenicafé study calls liming the most recommended practice for acid soils.12\n\nLime moves slowly. Spread on the surface of two Brazilian coffee plantations (1975–82), it acted mainly in the top 30 cm.6 A short trial in a Minas Gerais coffee plantation also found no correction of the deeper layer it sampled.8\n\nGypsum is calcium sulphate. It can neutralise aluminium without changing the pH, according to Cenicafé.2 It carries calcium deeper than lime. In a six-year Brazilian field trial with grain crops, gypsum raised calcium at every depth sampled, down to 1 m.7\n\n
\n
\n

Lime

\n
    \n
  1. Lime on the surface. Calcium and magnesium carbonates, spread over the soil.2,6
    2. \n
  2. Corrected topsoil. pH, calcium and magnesium rise and aluminium falls: mainly in the top 30 cm in a coffee trial, the top 10–20 cm in a grain-crop trial.6,7
    3. \n
  3. Acid subsoil. Not reached by surface lime; aluminium there restricts root growth.4,7,8
    4. \n
\n

Gypsum

\n
    \n
  1. Gypsum on the surface. Calcium sulphate; it does not change pH much.2,7
    2. \n
  2. Calcium moves down. Calcium rose at all depths to 1 m in a six-year grain-crop trial.7
    3. \n
  3. Subsoil with less aluminium saturation. At 40–60 cm it fell from 87% to below 20% in the same grain-crop trial.7
    4. \n
  4. Deeper roots. The aim given for high gypsum rates in Cerrado coffee, to ease water stress; not measured in coffee in the studies consulted here.9
    5. \n
\n
Where lime and gypsum act in the soil profile, from field trials in Brazil (lime in coffee and grain crops, gypsum movement in grain crops). Point at a number or a line of the legend to highlight the part. Schematic, not to scale.
\n
\n\n## How much lime\n\nThere is no universal dose. Cenicafé leaves it to the soil analysis, and a Brazilian handbook describes three calculation methods.2,13 One raises the share of bases on the soil particles, called base saturation, to a target.11,13 Coffee trials in Minas Gerais and Paraná took that target from their own state's manual.8,10\n\nIn Cenicafé's trials, 340 kg of agricultural lime per hectare and year raised production by about 15%.2 It gives a liming interval of one to two years.2\n\nMore is not better. In Brazilian trials, the lowest of three lime rates gave the best yields.6 The Deep dive gives the formulas and trial results.\n\n\n\n## Aluminium and manganese toxicity\n\n**Aluminium.** The 1982 greenhouse study grew arabica coffee for two years in solutions with 0, 0.074, 0.148, 0.296, 0.444 and 0.888 mmol/dm³ of aluminium.4 All growth measures fell significantly even at 0.074 mmol/dm³.4 The first symptom was slower root growth, with thicker roots and fewer laterals per unit of main root.4 Root tips stopped developing and turned brown.4\n\nLeaf symptoms came only after long exposure. Older leaves showed marginal chlorosis that spread to the centre of the blade.4 Visible injury went with leaf aluminium above 300 µg/g.4 Aluminium reduced the uptake of calcium, magnesium and phosphorus and raised the uptake of potassium and aluminium.4\n\nA companion study with seedlings held the solution at about pH 4.0 and varied aluminium and ionic strength.3 Growth followed the activity of the free Al³⁺ ion most closely. An Al³⁺ activity of 1.2 × 10⁻⁵ went with a 10% reduction of seedling growth.3 Only the abstract of this study was consulted.\n\n**Manganese.** A 1981 sand-culture study grew Catuaí Vermelho seedlings at rising manganese supply.5 Leaf manganese followed solution manganese closely.5 A yield reduction of about 25% went with 20 mg Mn/l in solution, or 2,500–2,800 µg Mn/g in the leaves.5 Excess manganese lowered leaf iron and calcium.5\n\n**Effects of lime on nutrient supply.** Liming does not simply make every nutrient more available. In the Paraná trials, lime raised exchangeable calcium and magnesium and the cation exchange capacity, but lowered exchangeable potassium.6 In the leaves it raised calcium and magnesium and lowered manganese, zinc and potassium.6 No study consulted for this entry measured a full curve of nutrient availability against pH in coffee soils, so no such chart is drawn here.\n\n## Nitrogen fertiliser and acidification\n\nCenicafé ranks nitrification among the most important acidifying processes.2 Its note lists the calcium carbonate needed to neutralise the acidity from one unit of nitrogen.2\n\n| Fertiliser | CaCO₃ equivalent per unit of N | Source |\n|---|---|---|\n| Urea | 3.6 | Cenicafé2 |\n| Ammonium nitrate | 3.6 | Cenicafé2 |\n| Diammonium phosphate (DAP) | 5.4 | Cenicafé2 |\n| Ammonium sulphate | 7.2 | Cenicafé2 |\n| Monoammonium phosphate (MAP) | 7.2 | Cenicafé2 |\n\nAt La Trinidad Experimental Station, Tolima, Colombia, four years of 300 kg N per hectare and year were applied as urea and ammonium nitrate.2 Soil pH fell by 0.27 units and exchangeable aluminium rose by 0.46 cmolc/kg.2 Aluminium saturation rose from 14% to 31%, and the effective cation exchange capacity fell by 25%.2 In a separate two-year study in 41 coffee lots, soil pH rose by 0.29 units on average when nitrogen applications (240 kg/ha a year) were stopped.2\n\n## How lime requirement is calculated\n\n**Base saturation method (Brazil).** The lime requirement in t/ha is NC = (V2 − V1) × T / PRNT.11,13 V1 is the base saturation measured in the soil test and V2 the target, both in %.13 T is the soil's cation exchange capacity.13 PRNT is the lime's relative neutralising power, in %.13\n\nThe target V2 is set by each state's manual. A coffee trial in Muzambinho, Minas Gerais, took it from the 1999 Minas Gerais recommendations, and a trial with young arabica plants in Paraná from the Paraná manual.8,10 The target values are not given here: the only sources consulted that state them are single Tier B papers.\n\n**Aluminium-based methods.** An Embrapa handbook describes three Brazilian methods: neutralising exchangeable aluminium while raising calcium and magnesium, raising base saturation, and a buffer solution.13 The aluminium method multiplies exchangeable aluminium by a factor and adds lime to lift calcium plus magnesium to a set level.13 The handbook's factors are written for beans and are not repeated here.\n\n**Colombia.** Cenicafé's 2016 note gives no formula; it leaves dose and source to the soil analysis.2 Its survey of soil-test records counts samples with pH ≤ 5 and with exchangeable Al³⁺ ≥ 1 (unit not stated in the table).2 In Antioquia, 81% of samples had Al³⁺ ≥ 1; in Valle del Cauca, 10% of 24,834 samples did.2 A 2008 Cenicafé study notes that KCl-extractable aluminium may overstate toxic aluminium in volcanic-ash soils (Andisols).12 It also states that in tropical soils, neutralising aluminium to a non-toxic level, around pH 5.5, is considered sufficient.12\n\nNo coffee institute outside Brazil and Colombia was consulted for this entry. Rates for Africa and Asia are therefore not given.\n\n## Lime materials and the calcium–magnesium balance\n\nLiming materials are mainly carbonates of calcium and/or magnesium; oxides, hydroxides and silicates are also used.2 Their composition varies. The dolomitic lime of the Paraná trial had 30.6% CaO and 20.1% MgO, with a PRNT of 95%.6 The agricultural lime in a Colombian nursery study had 46.7% CaO and 8.1% MgO.12\n\nIn that nursery study, five Colombian coffee soils received agricultural lime or Thomas slag in bags.12 Both raised pH and calcium in every soil and brought exchangeable aluminium below 1 cmol(+)/kg.12 High lime doses lowered magnesium in several soils, which the authors relate to competition between calcium and magnesium.12\n\nIn the Paraná trials, 10 t/ha of dolomitic lime raised the Ca:K ratio of the topsoil to 23.6:1 and 30.0:1 at the two sites in the first year sampled.6 The 2.5 t/ha rate gave 14.0:1 at both. The authors give about 13:1 as the target ratio.6\n\n## Field responses to lime\n\n**Paraná, Brazil, 1975–82.** Dolomitic lime was spread once, without incorporation, on adult plantations at Londrina (Bourbon Amarelo) and Florestópolis (Mundo Novo).6 Rates were 0, 2.5, 5 and 10 t/ha.6 Lime raised pH, cation exchange capacity, calcium and magnesium and lowered aluminium, mainly in the top 30 cm.6 Frost meant there was no harvest at Londrina in 1976 or at Florestópolis in 1982.6 At 2.5 t/ha the pH of that layer reached 5.0–5.6; at 10 t/ha it rose above 6.0.6\n\n
\n
\n
Total yield by lime rate at two sites in Paraná, Brazil: four harvests at Londrina (1975, 1977–79), two at Florestópolis (1980–81). The lowest lime rate, 2.5 t/ha, yielded most. Data: Chaves et al.6
\n
\n\nAt Londrina, the 2.5 t/ha plots gave 3,993 kg/ha of processed coffee over four harvests, against 2,719 kg/ha without lime.6 The 5 and 10 t/ha plots gave 2,879 and 2,936 kg/ha.6 That is 47% more with the lowest rate than without lime (our calculation).6 The authors link the best yields to lower aluminium and better cation ratios, and the lower yields at 10 t/ha to leaf potassium and zinc falling below critical levels.6\n\n**Colombia.** At the El Rosario (Antioquia) and San Antonio (Santander) experimental stations, 340 kg/ha a year of agricultural lime raised production by about 15% at both sites.2 Cenicafé reports that hydrated lime at up to 2 t/ha during establishment corrected acidity and improved growth.2 On a stony soil in Andes, Antioquia, 60 g of dolomitic lime per planting hole was enough to give thicker stems and longer branches.2 For acid soils, Cenicafé prefers large planting holes of 30 × 30 × 30 cm.2\n\n**Kodagu, India.** A 2016 study analysed a 2008 survey of 112 robusta estates at 800–1,000 m.1 Soil pH ranged from 4.5 to 6.3, with a mean of 5.5.1 Of the farmers, 63% added lime to reduce acidity and aluminium toxicity.1 Estates that limed harvested 20% more berries per infructescence (fruit cluster).1 Liming was also linked with 1.8 more flowers per inflorescence (5.4%) and less berry loss between June and September.1 The study recorded liming only as yes or no, without rates or soil changes.1\n\n**Young plants.** In a 2020 Paraná trial, IPR 107 arabica plants grew in soil columns with lime and gypsum from none to above the calculated requirement.10 Neither lime nor gypsum changed any measured growth trait.10\n\n## Gypsum and the subsoil\n\n**How it moves.** A trial at Jaguariaíva, Paraná, ran from 2002 to 2008 under no-till grain crops on a sandy loam Latosol.7 It combined 0–12 t/ha of gypsum with 0 or 3.42 t/ha of lime.7 Lime raised calcium only in the top 10 cm after 36 months and to 20 cm after 72 months.7 It raised the pH (CaCl₂) of the top 10 cm from 4.7 to 5.4 at 36 months.7\n\nGypsum raised calcium at every depth in proportion to the dose.7 It did not change subsoil aluminium concentration. Aluminium saturation at 40–60 cm still fell from 87% to just above 40% after 36 months and below 20% after 72 months.7 Gypsum raised pH below 60 cm by only 0.17 units at 12 t/ha.7 Without water stress, high gypsum doses lowered soybean yield, which the authors attribute to induced magnesium deficiency.7\n\n**Gypsum in coffee.** In the Brazilian Cerrado, a coffee study reports, farmers tend to apply gypsum above the recommended rate to improve the root environment in the deep subsoil and ease water stress.9 According to the Minas Gerais recommendations, the gypsum rate depends on soil texture, residual phosphorus or aluminium, calcium and magnesium.9\n\nA trial at São Roque de Minas, Minas Gerais, tested 0, 3.5, 7, 14 and 56 t/ha of gypsum on a very clayey red Latosol.9 Catucaí arabica had been planted in 2008 after 8 t/ha of dolomitic lime and 2 t/ha of gypsum.9\n\n
\n
\n
Cumulative yield of four harvests (2011–2014) by gypsum rate, São Roque de Minas, Minas Gerais, Brazil. Sack size is not stated in the paper. Data: Cogo et al.9
\n
\n\nThe 14 t/ha rate gave the highest cumulative yield, 300 sacks/ha against 238 without gypsum.9 In the first year, gypsum practically doubled yield, but this effect was not maintained in later years.9 The 56 t/ha rate gave 255 sacks/ha, less than 14 t/ha but more than no gypsum; the authors call it excessive.9 It also reduced the diversity of mycorrhizal fungi, while 7 t/ha increased it.9 No study consulted for this entry measured coffee root depth after gypsum.\n\n## Reference values\n\n| Parameter | Value | Context | Source |\n|---|---|---|---|\n| Suitable soil pH | 5.0–5.5 | Colombia, Cenicafé | Sadeghian2 |\n| Aluminium neutralised | Above pH 5.5 | Colombia, Cenicafé | Sadeghian2 |\n| Leaf aluminium with visible injury | Above 300 µg/g | Arabica, nutrient solution | Pavan & Bingham4 |\n| Al³⁺ activity for 10% less growth | 1.2 × 10⁻⁵ | Seedlings, pH ≈4.0 | Pavan & Bingham3 |\n| Leaf manganese with ≈25% yield loss | 2,500–2,800 µg/g | Catuaí Vermelho seedlings, sand culture | Pavan & Bingham5 |\n| Best lime rate | 2.5 t/ha dolomitic, once | Adult plantations, Paraná, 1975–82 | Chaves et al.6 |\n| Maintenance lime | 340 kg/ha a year; ≈15% more production | Two Cenicafé stations | Sadeghian2 |\n| Liming interval | 1–2 years | Colombia | Sadeghian2 |\n| Lime at planting | 60 g dolomitic per hole | Stony soil, Andes, Antioquia | Sadeghian2 |\n| Depth reached by surface lime | Mainly 0–30 cm | Coffee, Paraná, 1975–82 | Chaves et al.6 |\n| Acidification by N | pH −0.27; Al³⁺ +0.46 cmolc/kg | 300 kg N/ha a year for 4 years, Tolima | Sadeghian2 |\n| Gypsum, highest 4-year yield | 14 t/ha | Catucaí, Minas Gerais, 2011–2014 | Cogo et al.9 |\n| Gypsum rate the authors call excessive | 56 t/ha (255 vs 300 sacks/ha at 14 t/ha) | Same trial | Cogo et al.9 |\n| Liming and harvested berries | +20% per infructescence | 112 robusta estates, Kodagu, India | Boreux et al.1 |\n\n## Open questions\n\n- **Lime targets outside Brazil and Colombia.** No liming recommendation from African or Asian coffee research institutes was consulted for this entry.\n- **Gypsum and coffee roots.** The aim of deeper roots is stated for Cerrado coffee, but no study consulted here measured root depth or drought response after gypsum.9\n- **Yield response at the pH optimum.** The Colombian range of pH 5.0–5.5 comes from a technical note; no study consulted here gives a yield curve against soil pH for coffee.2\n- **Which aluminium measure predicts harm.** Solution studies point to Al³⁺ activity, while field practice uses KCl-extractable aluminium, which may overstate toxic aluminium in Andisols.3,12\n\n## Key studies\n\n
\n

Chaves, Pavan and Igue, 1984 – Response of coffee to lime. Spread 0–10 t/ha of dolomitic lime once on adult plantations at two sites in Paraná, Brazil, between 1975 and 1982. The effects stayed mainly in the top 30 cm; the lowest rate gave the best yields, and higher rates gave less.

\n

Pesquisa Agropecuária Brasileira 19(5), 573–582 · doi:10.1590/S1678-3921.pab1984.v19.15714

\n
\n\n
\n

Pavan and Bingham, 1982 – Aluminium toxicity in coffee. Grew arabica coffee for two years in nutrient solutions with six aluminium levels. Every growth measure fell even at the lowest level; roots thickened and lost laterals, and calcium, magnesium and phosphorus uptake fell.

\n

Pesquisa Agropecuária Brasileira 17(9), 1293–1302 · doi:10.1590/S1678-3921.pab1982.v17.16275

\n
\n\n
\n

Boreux et al., 2016 – Shade trees, irrigation and liming. Surveyed management and yield components on 112 robusta estates in Kodagu, India. Liming was linked with 20% more harvested berries per infructescence, though rates were not recorded.

\n

Agronomy for Sustainable Development 36, 42 · doi:10.1007/s13593-016-0377-7

\n
\n\n
\n

Pauletti et al., 2014 – Gypsum and lime over six years. Compared gypsum rates with and without lime in a no-till grain system in Paraná, Brazil, sampling to 1 m. Lime acted in the top 10–20 cm; gypsum raised calcium at all depths and lowered subsoil aluminium saturation.

\n

Revista Brasileira de Ciência do Solo 38(2), 495–505 · redalyc.org

\n
\n\n
\n

Cogo et al., 2020 – High gypsum rates in coffee. Tested 0–56 t/ha of gypsum in a Cerrado arabica plantation in Minas Gerais over four harvests. The 14 t/ha rate yielded most; 56 t/ha yielded less than 14 t/ha and reduced mycorrhizal diversity.

\n

Bragantia 79(4), 487–497 · doi:10.1590/1678-4499.20200014

\n
" }