Soil Testing and Fertility Management in Uganda

Direct answer

A useful soil sample starts with dividing the farm into zones that share similar soil type, slope and management history, then collecting multiple sub-samples per zone from a consistent depth, mixing them into one composite sample, and sending that composite - clearly labelled with location, depth and date - to a laboratory. There is no single universally agreed number of sub-samples or sampling depth; this article explains why reputable sources vary, gives the underlying decision rule, and tells you to confirm the exact protocol with whichever laboratory you use, since results are only comparable to that laboratory's interpretation bands if collected the way it specifies. Once results arrive, pH, organic matter and macronutrient levels should be read against general interpretation guidance - checked against visible field symptoms - to decide whether liming, organic-matter building, or a fertiliser purchase is the more urgent next step. Fertiliser application rates for specific crops are outside the scope of this guide and require a crop- and site-specific recommendation from an agronomist or laboratory.

Who this guide is for

Smallholder farmers deciding whether soil testing is worth the effort and cost; emerging-commercial farmers investing in a new plot or crop; farm managers responsible for fertility records across multiple fields; and investors who need to know whether a site's soil constraints have been properly assessed before capital is committed.

The problem being solved

Uganda's own soils-research history shows that this is not a new or hypothetical problem. A NARO-published review of a century of Ugandan soils research records that early administrators, observing lush vegetation around 1898, concluded the country's soils were simply "fertile," and that this conclusion - later shown to be an oversimplification - was repeated out of context for decades and contributed to poor attention to soil productivity management even after independence ([1], p.26). The same paper documents that national fertiliser consumption in Uganda has historically been low compared with several other countries in the region, and that use was concentrated on a small number of crops (tea, sugar, and in some periods cotton, tobacco and coffee) rather than spread across general smallholder cropping ([1], Table 3, p.27). Whatever the current picture looks like on any individual farm today, the historical pattern this paper documents is one of thin, uneven soil-fertility information - which is precisely the gap that a properly collected soil sample is meant to close.

Why sampling design matters more than the specific method used

Soil is remarkably variable, even within a single field. A composite sample - many individual cores mixed into one representative sample - exists specifically to capture that variability rather than being misled by it ([2], Section 3.1; [3], Section 2.3.1). Both manuals consulted for this article stress the same underlying point from different angles: a soil test is only meaningful if the sample was collected to represent the area it claims to describe, and sampling error is normally the dominant source of error in the whole testing process, not laboratory error ([2], Section 3.1).

Step-by-step: collecting a representative sample

1. Divide the field into sampling zones.

A composite sample should represent one area with broadly uniform soil, cropping and management history ([2], Section 3.1). Do not combine visibly different soil types, slopes, or cropping/fertiliser histories into one sample; sample each zone separately.

2. Choose a sampling pattern and avoid unrepresentative spots.

A zig-zag or diagonal pattern across the sampling area is recommended over ad hoc point selection ([2], Section 3.1; [3], Section 2.3.1, Figure 1). Deliberately avoid old channels, unevenly fertilised patches, areas near trees, old compost or manure piles, and other atypical spots - unless you are specifically investigating one of those spots as its own separate sample ([2], Section 3.1; [3], Section 2.3.1).

3. Collect individual cores.

There is no single universally accepted number of sub-samples for every field situation. One primary source notes that its own laboratory typically uses around eight sub-samples per hectare in a diagonal pattern, but that documented practice across different programmes ranges from roughly five to twenty-five borings per composite sample, over sampling units of about two to eight hectares - with fewer sub-samples appropriate where little or no fertiliser has been used, and more needed where fertility is more variable, for example under hand-broadcast fertiliser or mixed cropping-livestock use ([2], Section 3.1). The working decision rule from this evidence is: take more cores on larger or more variable fields, fewer on very small and uniform plots, and always ask your chosen laboratory what it expects, since laboratories calibrate their interpretation to their own stated collection method.

4. Sample to an appropriate, consistent depth.

Sources genuinely differ here, for a real reason. One primary manual states that for most general fertility purposes, soil is sampled to a depth of about 20 cm, since this is where most available phosphorus, nitrate-nitrogen and micronutrient concentrations relevant to crop growth are found - but that in some situations, particularly irrigated areas or where nitrate leaching is a concern, sampling to 60-100 cm is appropriate ([2], Section 3.1). A separate regional manual developed for African smallholder systems shows that different research bodies use different standard depth intervals for diagnostic purposes: for example, one system uses two depths (0-20 cm and 0-50 cm) for basic diagnostics, while a fuller research-grade soil profile may be sampled in several layers down to 150-180 cm ([3], Section 2.3.2, Table 1). These deeper, multi-layer protocols are designed for research-station soil characterisation, not routine farm fertility testing. The practical rule for a farm fertility test is: sample the depth relevant to the crop's main nutrient-uptake zone and to where you would incorporate lime or fertiliser - commonly around 20 cm for general fertility testing on cultivated land - and confirm the exact depth your chosen laboratory requires before you sample, since a result is only interpretable against that laboratory's own bands if collected to its specified depth.

5. Avoid sampling right after fertiliser or lime application.

Soil can be sampled at most times soil conditions permit, but sampling directly after a fertiliser or soil-amendment application should be avoided, since this will distort the reading rather than reflect the underlying soil status ([2], Section 3.1).

6. Composite and sub-sample.

Mix all individual cores from one zone thoroughly in a clean container, then reduce the mixed sample to a manageable size for submission using a "quartering" technique: spread the mixed soil out, divide it into quarters, and keep a representative portion ([2], Section 3.2; [3], Section 2.3.1, Figure 3).

7. Use appropriate sampling tools, especially for micronutrient testing.

An auger, spade or trowel can all be used; if no proper tool is available, a spade can be used to dig a small V-shaped hole and take a slice from the smooth vertical face ([2], Section 3.1). If the sample will be analysed for micronutrients such as zinc, use a stainless steel or otherwise ungalvanised tool and container - a galvanised coating is zinc oxide, and it will contaminate the reading ([2], Section 3.1, Section 3.2 - "technical remarks" on trace-element sampling).

8. Label and record.

Record, at minimum: date, location, zone/sample ID, depth sampled, and recent land use and input history. Sketch the field and mark where each sample was taken so future recommendations can be matched to the correct area ([2], Section 3.1).

9. Handle samples correctly, especially for nitrogen testing.

Soil intended for nitrate, nitrite or ammonium testing should ideally be analysed directly, without drying, because these forms of nitrogen are biologically active and change over time; if short-term storage is unavoidable, keep the sample refrigerated near 0°C and minimise the time between sampling and analysis ([2], Section 3.2). For other routine tests, air-drying is standard, but be aware that even air-drying is not neutral for every parameter - ammonium and nitrate content can still change with time and temperature during drying, and oven-drying at excessive temperature can affect nutrient availability more broadly, so air-drying (not oven-drying) is generally the recommended approach where laboratory guidance does not specify otherwise ([2], Section 3.2).

Choosing a laboratory or testing route

NARO's National Agricultural Research Laboratories at Kawanda have an institutional soil and plant analysis role: the institute's own website lists "Soil Testing Analysis" among its services, under a research programme mandated to generate tools and recommendations for soil and water management ([4]). Farmers should contact NARO directly to confirm current submission procedures, available tests, costs and turnaround times, since these operational details were not independently verifiable from the source reviewed for this article and can change over time. This article does not name or endorse any private laboratory or testing-kit provider; availability of such services should be verified directly and currently by the farmer.

Questions worth asking any laboratory or testing service before submitting a sample:

• What sampling depth and number of cores do you require for the composite sample?
• What extraction method do you use for pH - soil-to-water ratios of 1:1, 1:2.5 and 1:5 are all used in different contexts and are not directly comparable to one another ([3], Section 3.5.2) - and which does your reported result assume?
• What is your turnaround time, and do you provide interpretation guidance alongside raw figures?
• What is the total cost, including sample collection if that service is offered?

Interpreting the results: a general framework

The figures below are general soil-science interpretation guidance drawn from the manuals consulted for this article. They are a starting point for discussion with an agronomist or your testing laboratory - not a Uganda-specific regulatory threshold and not a substitute for your laboratory's own interpretation of its own results.

Soil pH. Suitable pH is not one fixed number: it depends on the crop, the soil type, the extraction method a laboratory used, the nutrient-management objective, and local agronomic conditions. Two of the sources reviewed for this section use different extraction methods and therefore report different-looking numeric cut-offs describing the same underlying chemistry - this is explained below rather than merged into one silent set of numbers.

One regional soil-testing manual, developed for Eastern and Southern African smallholder systems, presents the following pH bands, measured on a soil-to-water basis ([3], Section 3.5.2, Table 7, p.29 of the manual as paginated): below pH 5.5, soils are described as deficient in calcium and/or magnesium, with poor crop growth linked to low cation exchange capacity and possible aluminium toxicity and expected phosphorus deficiency; 5.5-6.5 is described as "satisfactory for most crops" while noting the soil should be monitored for further acidification; 6.5-7.5 is described as the "ideal range for most crops," with soil cation exchange capacity near full base saturation; 7.5-8.4 is described as containing free lime, giving good infiltration but reduced availability of phosphorus and micronutrients; and above 8.4 the manual states this "invariably indicates sodic soil," with poor physical soil conditions and possible root deterioration ([3], Table 7). The manual itself attributes this table to Estefan et al. (2013) and Juo & Franzluebbers (2003) - those two original studies were not independently opened for this article, so this table should be read as general guidance transmitted through the regional manual that was itself fully read, not as independently re-verified against the original publications.

As an independent cross-check (not African or Ugandan, but from a recognised university/government soil-science knowledge base, used here only to test whether the general pattern holds up, not to supply Uganda-specific numbers), a Western Australian government and university soil-science resource describes a broadly consistent pattern using a different extraction method - pH measured in a dilute calcium chloride (CaCl₂) solution rather than in water: below pH 5.5 (CaCl₂), it lists nutrient deficiencies (phosphorus, magnesium, calcium), reduced boron and molybdenum availability, and aluminium or manganese toxicity; above pH 8.3 (CaCl₂), it lists reduced micronutrient availability (notably zinc), boron toxicity, and sodicity-related structural problems ([6]). Critically, this second source explains why numeric pH cut-offs differ between soil-testing traditions: pH measured in water is typically 0.5-1.0 pH units higher than pH measured in a calcium-chloride solution for the same soil ([6]). This is a genuine, source-documented reason the exact numbers in different manuals will not match precisely - it is a methodological difference, not a disagreement about the underlying soil chemistry, and readers should always check which method their own laboratory used rather than assume all reported pH numbers are directly comparable.

Taken together, both independently-read sources agree on the same underlying pattern - acidic soils carry a real risk of aluminium/manganese toxicity and reduced availability of several major nutrients, near-neutral soils are described as most broadly suitable, and strongly alkaline/sodic soils carry a risk of micronutrient deficiency and poor soil structure - while disagreeing on the exact numeric cut-offs because they use different pH-measurement methods. None of these figures should be treated as a precise Ugandan threshold, a universal guarantee, or a rigid prescription; they are a starting point for a conversation with your laboratory or an agronomist, who can tell you which extraction method your own result used.

Crops sometimes described as more tolerant of acidic soil. The SIMLESA manual states that maize, cowpea, pigeon pea and coffee are "able to be cultivated successfully" on soils below pH 6.5 - soils it separately classifies as only mildly-to-moderately acidic, not "strongly acid" (its own strongly-acid threshold is below pH 5.0) ([3], Section 3.5, p.24). This is a claim about relative tolerance compared with more acid-sensitive crops, not a claim that any of these crops thrives in strongly acidic soil, and it is a single-source claim for maize, cowpea and pigeon pea specifically - independent verification of those three crops was attempted for this article (several peer-reviewed studies on maize and aluminium-toxicity tolerance in tropical acid soils were located) but could not be completed to full-text standard within this pass, since the most relevant peer-reviewed review article could not be opened due to a website access restriction. That limitation is disclosed here rather than used as a reason to delete the claim, consistent with the instruction to keep and qualify useful information rather than remove it when reasonable verification has been attempted.

For coffee specifically, a Uganda-specific official source could be verified directly: a Ministry of Agriculture, Animal Industry and Fisheries (MAAIF) training manual for coffee production in Uganda states that Arabica coffee requires "deep, well drained, fertile and slightly acidic loamy soils (of pH 4.5 to 5)" ([5]) - a stated pH range claim, not merely a "tolerates acidic conditions" claim, and narrower/more specific than SIMLESA's general "pH < 6.5" statement. The same Ugandan manual does not give a specific pH figure for Robusta coffee, describing its soil requirement only as "deep, well drained fertile loamy soils" without a stated pH band ([5]) - so the SIMLESA manual's broader "coffee" claim and this Ugandan source's Arabica-specific figure should be read as describing two different things (a general tolerance statement versus a stated cultivation range for one coffee type), not silently combined into one number.

In every case: varietal differences within a crop, aluminium and manganese toxicity risk, the availability of other nutrients, drainage, and other local soil conditions can all still affect whether a given field performs well at a given pH, even within a "tolerant" crop's usual range. None of these crops should be described as universally acid-tolerant, and a favourable pH reading does not by itself confirm a field is suitable for any of them.

Organic matter. Organic matter contributes to nutrient-holding capacity, water retention and soil structure. Darker topsoil colour is generally, though only approximately, associated with higher organic matter content; formal colour-based estimation charts exist for trained use, but a laboratory organic-carbon or organic-matter percentage is more reliable for decision-making than a visual estimate ([3], Section 3.2).

Macronutrients and micronutrients. These are typically reported by laboratories against their own "low/medium/high" or "deficient/adequate" bands rather than one universal number, because the same raw figure can mean different things depending on soil type, crop and extraction method used.

Reading results alongside field symptoms. A laboratory number should be checked against what is visible in the field - stunted or pale growth, poor root development, patchy performance, or persistent waterlogging - since these can point to problems (compaction, drainage, disease, pests) that a standard nutrient panel does not detect on its own.

A note on the international and regional evidence used in this article

Two of the three primary technical sources behind the sampling and interpretation guidance in this article were not developed specifically for Uganda. The ICARDA laboratory manual [2] was developed for the West Asia and North Africa region, an arid-to-semi-arid zone with different rainfall and soil-forming conditions than most of Uganda; its sampling-procedure principles (composite sampling logic, avoiding unrepresentative spots, depth reasoning, contamination avoidance) are general soil-science practice and reasonably transferable, but its specific interpretation values were developed and calibrated for its own region and should not be assumed to transfer precisely to Ugandan soils without local confirmation. The SIMLESA manual [3] was developed specifically for smallholder farming systems in Eastern and Southern Africa under an Australian government-funded (ACIAR) programme delivered through CIMMYT, with field validation at agricultural research stations in Ethiopia, Mozambique and Malawi - regionally closer to Uganda, but still not Uganda-specific, and its authors themselves note that the manual's rapid field methods are not intended to replace full laboratory analysis where that is available. The one Uganda-specific source used, a NARO-published historical review [1], provides genuine Ugandan evidence but is a historical account published in 2000, covering research mostly conducted before that date; it should be read as historical context for why soil testing matters in Uganda, not as a current statement of any individual farm's soil status.

Local Ugandan considerations

The 2000 NARO review documents that Uganda's national reconnaissance soil survey (completed in the late 1950s) found soil reaction (pH) across the country spanning the very wide range of pH 1.4 to 9.0 across 138 distinct soil-mapping units, and that soil types vary regionally - the same paper's fertiliser-trial records, for example, distinguish trials conducted on the volcanic-ash-derived soils of the Mount Elgon area from trials elsewhere in central and western Uganda, and separately note that communal grazing land in areas such as Karamoja was recorded as being particularly susceptible to erosion ([1], p.26-27, Table 1). This is strong historical evidence that a single national fertility or soil-type picture does not apply uniformly across Uganda, even though it cannot tell you what any specific field looks like today. Laboratory access has also historically been concentrated at a small number of institutions - the same source records that Uganda's national soils laboratory has operated from Kawanda since the 1930s, with Makerere University's soil science department as the other major concentration of soil-testing and analytical capacity in the country ([1], p.24) - which is a relevant starting point when deciding where to send a sample, alongside directly confirming current options with NARO.

Measurements and records to keep

• A dated sample log per zone (location, depth, core count where known, sampler name).
• The full laboratory report, retained rather than only a summary interpretation.
• A simple field map showing which zone each sample represents.
• A record of any lime, manure, compost or mineral fertiliser applied after each test, with approximate quantity and date.
• Visual field observations (growth patterns, waterlogging, erosion) logged at the same time as sampling.

How often to retest - a proposed decision framework

This is an editorial, practical decision framework rather than a claim drawn from a cited study: there is no single universally correct testing interval, and the right frequency depends on management intensity, crop value, and how much land use is changing. As a working rule: retest before establishing a new high-value crop or greenhouse system; retest a zone after a significant management change (new irrigation, heavy liming, a shift to more intensive cropping); and for actively managed commercial land, consider a routine retest on a multi-year cycle to track trends rather than treating one test as permanently valid. A historical test does not reflect current conditions on land that has since been cropped intensively, limed, or heavily fertilised.

Common mistakes

• Submitting one grab sample instead of a composite of multiple cores.
• Mixing distinct soil zones (different slope, colour, or cropping history) into a single sample.
• Sampling immediately beside a fence line, path, or recently fertilised patch and treating it as representative of the whole field.
• Sampling at an inconsistent or unrecorded depth, making it impossible to compare results over time.
• Sampling directly after a fertiliser or lime application.
• Using galvanised tools or containers for micronutrient sampling.
• Treating a single laboratory number as a fertiliser prescription without asking what it means for the specific crop intended.
• Not asking the laboratory what sampling protocol its interpretation bands assume.

Limitations and exceptions

This guide can help a farmer collect a defensible soil sample and understand, in general terms, what common results mean. It cannot:

• Diagnose a specific field's problem - that requires laboratory results interpreted by a qualified agronomist or soil scientist familiar with the crop and locality.
• Provide fertiliser application rates - these depend on the crop, target yield, soil test result, and local product formulations, and are outside the scope of this article.
• Guarantee laboratory accuracy, cost or turnaround - this depends entirely on the specific laboratory used and must be confirmed directly with that laboratory.
• Substitute for a site visit where drainage, compaction, salinity or contamination is suspected.
• Confirm current government regulations, subsidy schemes or input-quality standards - these should be confirmed directly with MAAIF, UNBS or the relevant current authority.

When professional support is required

• An agronomist or soil scientist should be consulted to translate a laboratory report into a specific fertiliser or lime plan for a named crop and target yield.
• The testing laboratory itself should be the first point of contact for questions about its specific methods, units, and turnaround.
• A drainage or irrigation engineer should be consulted where waterlogging, salinity, or irrigation water quality is suspected as a contributing factor.
• NARO should be contacted directly for the most current guidance on available testing services, since availability and operational details were not exhaustively verifiable at the time of writing.

What the reader should do next

1. Walk the farm and mark out distinct sampling zones based on visible soil, slope and management differences.
2. Identify a candidate laboratory or testing service and ask the questions listed above before sampling.
3. Collect a properly composited sample per zone, following the depth and core-count guidance the chosen laboratory specifies, and avoid sampling right after a fertiliser or lime application.
4. Keep a written sampling record alongside the submitted sample.
5. On receiving results, discuss them with an agronomist or NARO contact before making a fertiliser or lime purchase decision.
6. Build a simple retesting schedule based on the decision framework above, rather than treating one test as permanent.