Coffee soils and plant nutrition: what five studies found

What the soil is asked to do

Sourced

The physiology review behind the guide to the coffee plant places the main part of an arabica tree's root system in roughly the top 30 centimetres of soil, in a circle about a metre and a half across. A plant rooted that shallowly depends on the surface layer for water as well as nutrients, which is the layer most exposed to drying, to erosion and to whatever is put on it.

Coffee is also a perennial. An annual crop meets a field for a season; a coffee tree stays for many years, and the next planting usually goes into the same ground. Problems that an annual crop escapes by rotation accumulate under coffee.

Nitrogen appears to matter enough to have left a mark in the genome. The 2024 study of arabica's genome described in the guide to coffee genetics looked for genes that differ most between cultivated and wild trees. The set was enriched for genes involved in the response to nitrogen starvation and included relatives of ammonium transporters. That is a statistical signal in DNA and its authors do not claim more for it.

What decades of coffee do to the soil under it

Sourced

A study in Hainan, in southern China, compared the soil of four coffee fields that had been planted to coffee continuously for 4, 18, 26 and 57 years. Acidity did not simply rise with age: the soil's pH fell from 6.3 in the youngest field to 4.5 in the field planted for 26 years, and in the oldest field it was 5.4, about where the 18-year field stood. Electrical conductivity and available phosphorus, iron and zinc rose. Organic matter was highest in the youngest field. The number of kinds of bacteria and fungi detected fell significantly after 26 years.

The authors then grew coffee seedlings in pots of soil from each field. Seedlings grew less in soil from older fields, with the lowest shoot and root weight in soil from the 57-year field. That step matters, because it shows the soil itself had changed in a way that held coffee back, and was not only keeping company with old trees.

What it cannot show is why. Four fields are four places as well as four ages, and they will have been fertilised differently for different lengths of time; the study measured what the soils had become, not what made them so. The authors describe the pattern as a continuous-cropping obstacle, a term used for several crops, and attribute it to the combination of chemical change and a shifted microbial community. It is one study in one region.

Nitrogen, and what one field trial found

Sourced

Nitrogen is the nutrient coffee growers spend most on, and the one whose excess has costs beyond the farm: fertiliser nitrogen that is not taken up can leave the soil as nitrous oxide, a greenhouse gas, and its transformations in the soil acidify it. A trial published in 2025 by researchers at the University of Costa Rica, with agronomists from the farm's owner, a coffee company, tested this on an Andisol, a soil formed from volcanic ash, in Alajuela.

Three ways of supplying nitrogen were compared on eighteen plots spread across two parts of the farm, over one crop cycle of eight months: one common form of nitrogen fertiliser, the same form with an additive that slows its conversion in the soil, and a formula based on a different form. Yield did not differ significantly between the three. It differed strongly between the two parts of the farm, which the experimenters had separated by the lie of the land: one gave roughly 5,100 to 7,300 kilograms of cherries per hectare and the other 2,500 to 3,000. The soil of the better sector was a loam with twice the clay of the other's sandy loam, 20 per cent against 10 on the paper's table, and its available phosphorus roughly tripled during the season while the other's did not change. At the start the two had tested alike in everything but iron. Nitrous oxide emissions over the season were highest under the first of the three, the treatment that also received the most nitrogen. The soils took up methane.

The authors' own title states the finding: yield was influenced by soil properties, not by nitrogen fertilisation strategy. One season on one farm does not show that nitrogen strategy never matters, and every plot received nitrogen: there was no unfertilised control. Nor does the comparison of sectors show which property of the soil made the difference, since texture, phosphorus and position on the slope went together; the authors offer the phosphorus as a partial explanation. What it shows is that, there, differences in the soil within a single farm outweighed the differences between the fertiliser programmes being compared. The amounts applied are in the paper and are not repeated here.

Phosphorus, and a difference between the species

Sourced

Phosphorus is the nutrient tropical soils are worst at giving up. A Brazilian study of phosphorus use explains why in its introduction: the old, heavily weathered soils where most Brazilian coffee grows are low in phosphate a plant can use, and phosphate added as fertiliser is held by clay minerals or locked up by iron and aluminium.

The study, read here in its summary and introduction, grew 21 arabica cultivars and four robusta cultivars at low and high phosphate. Growth fell at low phosphate in all of them. The two species coped differently. The robusta cultivars had more root relative to shoot and were significantly more efficient at taking phosphorus up; the arabica cultivars were more efficient at using what they had taken up to make plant. Young leaves held more phosphorus than any other tissue.

This was an experiment under controlled phosphate supply, and its purpose was to find traits for breeding. It does not say how either species yields on a poor soil, and the cultivars of each species differed among themselves.

The fungi that live in coffee roots

Sourced

Coffee roots are colonised by arbuscular mycorrhizal fungi, which extend into the soil and pass nutrients, phosphorus above all, to the plant in exchange for sugar. A Brazilian study at Araponga, in Minas Gerais, compared these fungi at three localities, each with a plot of coffee managed conventionally, a plot managed on agroecological lines and a nearby fragment of forest, at flowering, grain filling and harvest. Agroecological here meant less fertiliser, green manures and other crops or trees among the coffee; those plots were still fertilised.

Counting kinds of spore under a microscope showed no difference between the two ways of growing coffee. Nor did the number of spores, or the share of root the fungi had colonised, which was lower under both kinds of coffee than in the forest. A molecular fingerprint of the community did: by that measure the agroecological coffee resembled the forest and was more diverse than the conventional coffee. Place and season also changed which fungi were found. The authors make a point of the disagreement between their two methods, and it is the most transferable thing in the paper: how diverse a soil's fungi appear depends on how they are counted.

The study measured diversity. It did not measure whether the more diverse fungal community fed the trees better or changed the crop.

Shade trees that fix nitrogen

Sourced

Many of the trees planted over coffee are legumes, which house bacteria that turn nitrogen from the air into a form plants can use. A field experiment in the Ecuadorian Amazon grew five-year-old robusta under four shade arrangements and in full sun, crossed with conventional and organic management at two intensities.

Robusta grown under either of two leguminous trees was about a tenth taller than robusta in full sun, and its leaves held a quarter to a third more chlorophyll. The authors treat chlorophyll as a measure of leaf nitrogen, having calibrated their meter against analysed leaves in seventeen classes of reading, three leaves to a class, and their summary puts the difference in nitrogen at about a fifth. Those two were also the plots with the most shade, about a quarter against under a tenth in the other shaded plots, so the experiment cannot tell the trees' nitrogen from their shade. Cherry yield over three months of picking did not differ significantly between the shade arrangements, and was higher under conventional than under organic management. Under organic management the leaves of coffee in full sun or light shade held less chlorophyll than under conventional management; under the two legume canopies that difference disappeared. The authors found no correlation between height, yield and leaf chlorophyll, and say plainly that longer study is needed: three monthly assessments of young shrubs are a beginning.

So the sign of nitrogen was in the leaves, and the experiment does not show it in the harvest. That distinction, between a nutrient measured in the plant and a yield or a cup measured at the end, runs through this subject.

Depth, drainage, acidity and slope

Sourced

Less was found on the physical side of coffee soils than on their chemistry, and what follows rests mainly on one growers' manual, published by the FAO for arabica in Laos, with measurements from two of the studies above. It is a description of what such documents ask of a soil, not a survey of coffee soils.

The manual asks first for drainage and depth: a free-draining soil at least a metre deep, because coffee, in its words, will not tolerate waterlogging. It names a fertile red earth of volcanic origin or a deep sandy loam as the kinds it prefers and heavy clay or poorly draining ground as the kinds to avoid. That is a statement about water and air around the roots. It says nothing about flavour, and the manual does not claim that it does.

Texture and density can be measured, and the Costa Rican nitrogen trial shows how much they vary within one farm. Its two sectors were a loam and a sandy loam, with 20 and 10 per cent clay, and both were light, about three quarters of a gram per cubic centimetre, which its authors describe as typical of soils formed from volcanic ash. The sector with more clay and silt yielded about twice as much. Volcanic, in other words, names where a soil came from and not how it behaves.

Acidity is the property the manual returns to most. It gives a range of soil acidity that coffee prefers, says many cultivated soils in its region are more acid than that, and explains why it matters: at low pH several nutrients become less available to the plant. The Brazilian study of root fungi shows the same thing from the other side. Its forest soils, which were not fertilised, had a pH near 4.3 and about four fifths of their exchange capacity occupied by aluminium; the coffee plots beside them, under either kind of management, were near pH 5 with less aluminium, in most cases far less, and several times the calcium and magnesium. The authors put the coffee soils' chemistry down to fertilisation. That is the reverse of the Hainan result, and a reminder that what growing coffee does to a soil depends on what is put on it.

On a slope the concern is the soil staying where it is. The manual has rows run across the slope with a slight fall for drainage on gentler ground, planted along the contour on steeper ground, with a ground cover between the trees against erosion and a mulch under them to hold moisture. It also wants land prepared and shade trees established a year before the coffee arrives. No study that measured erosion or compaction under coffee was opened, though an index search shows that plot measurements of runoff and soil loss under coffee have been published, so nothing is said here about how much soil is lost under any system, or about what machinery or foot traffic does to it.

The other nutrients, and what each is for

Sourced

A plant is mostly carbon, hydrogen and oxygen, which it takes from air and water. The FAO manual lists thirteen further elements it takes from the soil and what each is chiefly used for. The roles are general plant physiology as the manual summarises it. The larger and smaller groups are named for the amounts a plant needs, not for importance.

The manual illustrates the old rule that growth is limited by whichever nutrient is shortest in supply with a barrel whose water can rise no higher than its shortest stave: supplying more of anything else changes nothing until that one is made good. It is a teaching picture and a useful one. It is also why a table like this cannot tell anyone what a particular soil lacks.

What a harvest takes away

Sourced

A forest recycles its nutrients; a farm exports them in the crop. The FAO manual quotes a figure from India for the size of the export: for each tonne of green coffee, about six tonnes of ripe fruit, roughly 40 kilograms of nitrogen, 2.2 of phosphorus and 53 of potassium, which the manual words as what must be replaced each year and not as a measurement of what left. It also gives a table of what a hectare of trees yielding a tonne of green coffee takes up in a year, by part of the tree. Of 112 kilograms of nitrogen and 125 of potassium, the fruit accounts for 30 and 35. The leaves hold more than the fruit, 53 and 45, and the manual's point is that the leaves' share returns to the soil when they fall and the fruit's does not. Neither figure is sourced in the manual beyond the country, and they are given here for their proportions.

A study of sixteen robusta clones in the Brazilian Amazon measured the same thing directly, by drying ripe fruit, separating seed from husk and analysing each. Its authors report an order resembling an earlier study's: in the seed nitrogen first, then potassium, phosphorus and calcium; in the husk potassium first. Clones differed, both in what they accumulated and in how much of the dried fruit was seed, from about half to about seven tenths. Its authors draw two conclusions that reach beyond their site: that the nutrients a crop takes up in quantity are nitrogen, potassium, calcium and phosphorus, and that husk returned to the field puts part of them back.

Removal is an accounting figure. It says what leaves, not what has to be applied, since soils differ in what they hold and release and since much of what is applied is lost before a root reaches it; the Brazilian phosphorus study and the Costa Rican trial above are both about that gap. The amounts either source would have a grower apply are in the documents and are not repeated here.

Leaf and soil analysis: what a sample is

Sourced

Advice about feeding coffee, where it is any good, starts from two laboratory tests, and the FAO manual explains what each sample has to be. A soil sample is an average of a block: soil from the top fifteen centimetres at twenty or more points, mixed, with the litter brushed aside and nothing taken beside a shade tree or just after fertiliser has gone on. A leaf sample is an average too: the third or fourth pair of leaves from the tip of an actively growing branch, from dozens of ordinary trees, neither the sickest nor the best, at the same stage each year. The manual prefers the time before flowering for both.

The two tests answer different questions. A soil test says what is in the ground in a form a particular chemical extraction can remove. A leaf test says what the tree has managed to take up. The manual says of nitrogen that the leaf test is the more relevant, and the legume shade experiment above is an example of a leaf measurement moving while the yield did not.

The manual prints tables of the levels it calls optimum for leaves and for soils, drawn, it says, from plantations around the world, and attaches a note that deserves more attention than the tables: different extraction methods give different results and different optimum levels. A soil figure means little without the method that produced it, and a table from one laboratory cannot be laid over a report from another. For that reason, and because a range of that kind is a prescription once it is on a page, none of those levels is reproduced here.

Before any laboratory there are the leaves themselves. The manual sorts the signs of deficiency by where they first appear. Shortages of phosphorus, potassium and magnesium show first in older leaves, as yellowing between the veins or dead tissue creeping in from the edge; shortages of iron, zinc, calcium and boron show first in the youngest leaves and shoot tips. Plant physiology has a standard reason for the pattern, which the manual does not spell out: some nutrients can be moved from old tissue to new and others cannot. The manual also ties shortage of nitrogen and potassium to the die-back of heavily cropped, unshaded trees described in the guide to the coffee plant. The manual says each nutrient has its own symptoms and shows them in photographs. CoffeeHQ's caution, which is not the manual's, is that symptoms overlap with each other and with disease, and are a prompt to test and not a diagnosis.

Soil and flavour: what is claimed and what was found

CoffeeHQ explanation

Statements a reader will meet on bags and in origin descriptions, set against what was opened for this page. The verdicts are CoffeeHQ's, and they say what was read and not what is known: an index search returns studies that set soil measurements beside cup scores, several of them from Ethiopia's coffee forests, and none of them was read.

Volcanic soil makes better coffee

A trade expectation. Nothing read for this page compares the cup from volcanic and non-volcanic soils. The one trial on a volcanic soil found yield varying twofold between two parts of the same farm, which is a reminder that volcanic is not one soil.

Minerals in the soil can be tasted in the cup

Unsupported by anything read. No study opened here traces a mineral from soil to seed to flavour, and the guide to green coffee chemistry gives no account of minerals in the seed because none was read.

Rich soil means sweet coffee

Unsupported by anything read. The guide to green coffee chemistry notes that no study it drew on pairs sugar in the green seed with sweetness in the cup, and no study read here pairs a soil measurement with either.

Coffee exhausts its soil

Supported in one respect by one study: under continuous coffee the Hainan soils became more acid and seedlings grew worse in the older ones. Exhausts is too simple a word, since several nutrients had accumulated.

Organic or agroecological management gives healthier soil

Supported for one measure in one study, the diversity of root fungi by a molecular method, and not by the other measure in the same study. No effect on the crop was measured.

What a reader will not find about soil here

CoffeeHQ explanation

No fertiliser, amount, timing or target for any soil or leaf measurement is given, and the amounts used in the trials and the manual are deliberately left out. Potassium, calcium, magnesium and the trace elements are described by their roles and by what a harvest removes; no trial of any of them in coffee was opened. Soil depth, drainage, acidity and slope are described from one manual. No measurement of erosion or compaction under coffee was read, and composting and liming are not described as practices. Every study here is one place over a short period, and most of them measured the soil or the plant and not the harvest or the cup.

What to do next

• Managing a coffee farm — Shade, spacing, water and pruning, each as a trade the sources describe better than they resolve.
• The coffee plant — The shallow roots, the fruit's long development and the seed this soil has to feed.
• Coffee pests and diseases — Root-knot nematodes and coffee wilt, the two problems that live in the soil.
• Coffee genetics and breeding — Where the genome study that found a nitrogen signal comes from, and what else it found.