A laboratory writes its report for someone who already knows how to read it, and copying that report is a different job from interpreting it. What turns a paid analysis into a farm record is one page per field carrying six things: the sampling date, the depth that was sampled, the name of the laboratory, the name of the method printed beside each determination, the results table with the number and the unit exactly as the report prints them, and a line at the foot with the name of whoever read it and the date. Twenty minutes per field, with the file open and a sheet beside it.
The report arrived as an attachment and got opened on a phone in the middle of something else. Whoever pulled the cores was not whoever opened the file, whoever opened the file was not whoever decided anything, and nobody wrote a line down. Six months later somebody asks how much phosphorus the back field had, and answering means searching an inbox for a word nobody remembers. None of that depends on how many fields there are. The analysis itself was never lost. What the farm cannot produce is a copy of it.
What the loss costs shows up twice. Once when a decision gets taken off the letter beside the number instead of the number, by somebody who did not notice that last time the samples went to a different laboratory. And once when a buyer asks for the report on the block he is about to take over, and the farm sends the whole file, every field in it, because pulling one out looked like too much work that afternoon.
What does every report carry, whatever its layout?
Enough to build a nutrient recommendation on, and the same items every time. Ohio State University Extension opens its guide to reading these documents by conceding the obvious, that “Soil test report formats vary from laboratory to laboratory”, and then states the half that makes a fixed page possible: “any standard soil test report has the critical information needed to make a nutrient recommendation”.
The layout changes with the laboratory. The items you copy out of it do not. That is why one page with the same six fields works for every field on the farm and for every laboratory the farm has ever used, and why a page that leaves a field blank is reporting something real rather than being incomplete.
Why copy the number and never the letter?
The letter is the laboratory’s ruler, not a property of your ground. University of Minnesota Extension puts it without hedging: “A soil testing laboratory can use any range of values that it chooses”, and names the consequence in the next sentence, that “A difference in the range of values for each relative level is one source of confusion that adds to the difficulty of evaluating results from more than one soil testing laboratory”. Change laboratory between one test and the next and the letter everybody remembers can begin at a different number, on a scale that laboratory chose. The number holds still. The letter does not.
The same page carries the harder point, which is that the number beside each nutrient is an index value, and is neither the amount available to the crop nor the total amount present in the soil. Nobody on the farm is going to reconstruct that from memory either. Copy it as printed and leave the reading to whoever signs the recommendation.
Why does the unit travel with the number?
Because two reports can measure the same soil and disagree by a factor of two on paper. Minnesota Extension gives the arithmetic used in the United States, parts per million multiplied by two gives pounds per acre, and then warns the reader “to be aware of the difference between ppm and pounds per acre when reading the units on the soil test report”.
Ohio State names the same trap next to its twin, saying that “The units and the test used for each nutrient’s reported soil test value are two areas of concern”. Copy the unit that is printed, in the same column as the number it belongs to, and do not translate it into the unit you are used to. The translation is obvious only while you still remember which report it came out of.
Why is depth a field and not a footnote?
The number is a number for a layer, and the layer was chosen before the sample left the field. The Embrapa sampling manual, written for Brazilian conditions, sends the sample for annual crops to the tilled layer, 0 cm to 20 cm, and asks for two layers on pasture and perennial crops, 0 cm to 20 cm and 20 cm to 40 cm.
Minnesota Extension shows the same dependence from the other end, converting nitrate concentration into pounds per acre for each increment of depth that was sampled, which means the depth is inside the number before anybody reads it. Set a pasture page beside a cropped field page without the depth written on both and you are comparing two different slices of ground while believing you are comparing two fields.
What does the laboratory’s name buy you?
A name you can ask a question of, and a way to find out whether anyone measures that laboratory at all. Mazzoni and colleagues, writing in Applied Sciences in 2026 about proficiency tests run on 27 laboratories in Belgium, France, Germany and Morocco, record that laboratories work under recognized quality systems “such as accreditation or approval by Ministries of Agriculture”, and that taking part in those tests “is also a mandatory requirement for obtaining accreditation under the ISO/IEC 17025 standard”. Those laboratories put themselves forward to have their own performance measured, so what the paper describes is what a laboratory willing to be measured achieves, and not an average of the world’s laboratories.
The same arrangement turns up on the other side of the Atlantic and in the southern half of it. The Embrapa manual closes with a list of soil and plant analysis laboratories authorized by the Ministry of Agriculture to provide the service in Brazil, and INTA, writing about the soil laboratory at its Catamarca station in Argentina, records that those analyses run on technical protocols and on the IRAM-SAGPyA standards, recommended by the laboratory network the station belongs to. Three jurisdictions, one arrangement: somebody recognizes the laboratories and somebody names the methods. Your report shows the result of both and says neither, which is exactly what the two fields on your page are for.
What a seal of quality does not tell you
That every determination on your report was inside the standard. Griebeler and colleagues examined the quality control of the ROLAS-RS/SC network in southern Brazil and published the result in Revista Ceres in 2016, and what they found at the lower edge of the seal is worth carrying around. The grade that earns the seal rests on the yearly average of every attribute measured in the soil analysis, so a determination whose precision falls short can end up hidden by the dilution effect of the determinations that came out precise, and the way the grade is calculated hides whether the weak determinations are spread across every attribute or piled onto a single one.
Then they scored that same year attribute by attribute instead of averaging it, and the seal came apart. Most of the laboratories that had earned it turned out to be carrying at least one determination below the precision the network requires, and some of the ones holding the top grade were carrying a determination at the bottom grade. What the authors wrote into their conclusions is the plain version of it: the yearly average does not reveal which determinations fell short of the minimum.
That is one network, one year of records, and the authors were auditing the network’s own scoring rather than testing whether the laboratories tell the truth. It is still the reason the method goes on the page beside each determination instead of once at the top, and the reason the laboratory’s name is a field and not a compliment.
Why the method name sits beside each determination
Because the number belongs to the method, and the method has a name that is already printed on your report. Ohio State records that most soil testing laboratories use Mehlich-3 as the default extractant, then publishes multipliers that move a Bray P1 phosphorus value or an ammonium acetate potassium value onto the Mehlich-3 scale, tied to the recommendation tables of three American states.
A published multiplier is the proof that two tests do not report the same number for the same soil. Copying the method name costs one line, and it is what lets somebody decide later whether two numbers from the same field can be set side by side at all. Doing the conversion yourself belongs to whoever signs the recommendation, under the tables of the place where they sign it.
Who read it, and on what date?
Nobody on the farm owns the result until a name sits under it. UMass Extension breaks soil testing into four steps and hands them out, stating that “The first step is the responsibility of the farmer while the next three are performed by a soil testing lab”. Collecting is the farm’s. Analyzing the sample, interpreting the result and writing the recommendation are done by people who do not work there and go home with the reasoning.
The United States conservation standard for nutrient management writes the missing line straight into the record. The New York version of the USDA Natural Resources Conservation Service practice standard 590 requires the operator to “Maintain records for at least 5 years to document plan implementation and maintenance”, opens the required list with “All test results (soil, water, compost, manure, organic by-product, and plant tissue sample analyses) upon which the nutrient management plan is based”, and further down the same list asks for “Dates of plan review, name of reviewer, and recommended adjustments resulting from the review”.
A reader outside the United States goes looking for the equivalent clause in whichever scheme audits their farm, and finds the same two items sitting together: the results, and the name of the person who reviewed them.
What goes in the field you cannot fill today?
A name and a date, never a guess. Ohio State closes its discussion of extractants by telling the reader to contact the laboratory to understand the values shown in its report whenever the extractant is one the fact sheet does not name, which converts a blank into a phone call with an owner and a deadline.
The laboratory can answer because the farm already gave it the context: the Embrapa manual reproduces the form that goes with the sample, asking for the owner, the property, the municipality, the sender, the sample number, the date and the number of hectares the sample represents, and says the form exists to guide the laboratory in interpreting the data and in recommending liming and fertilizer. Somebody on the farm filled that in. Somebody on the farm can ring up and ask what came back.
What the six fields do
| Field on the page | Copied from | What breaks when it is blank |
|---|---|---|
| Sampling date | The header of the report | Nothing can be called current or overdue, and two reports cannot be ordered |
| Depth sampled | The header, beside the sample identification | Pasture and cropped ground get compared across different slices of soil |
| Laboratory | The header or the letterhead | No one to ask, and no way to check who recognizes that laboratory |
| Method beside each determination | The results table, next to the nutrient | Two numbers on the same field cannot be judged comparable or not |
| Number and unit as printed | The results table, copied without conversion | A value in one unit gets read against a range written in another |
| Name of the reader and the date | Written by hand, by the person who read it | The result belongs to the laboratory, and nobody on the farm answers for it |
Which fields have a current test at all, and which are overdue, is a separate sheet with a declared sampling interval at the top of it, and that sheet is the queue this one works through. Three of these pages on the same field, in three different years, become the field’s soil test series, and the series only holds if the depth, the laboratory and the method stayed constant or the change was written down.
The name at the foot of the page is where the trail starts when somebody asks the soil test the recommendation rests on for a product already spread on that ground. All of it is the control function doing its work, which is one of the four functions of management, and the rest of the production axis is built the same way.
Where to start
Twenty minutes per field, with the report open on one side and a blank sheet on the other. Do the field somebody asked about most recently first.
The page you end up with is worth less than the list of blanks beside it. Blanks are what the farm can still fix, by making one call to a laboratory that has the answer written down, and a page with three of them and a name against each is a working document. A page with six fields full and no idea which of them came off the report and which came off somebody’s recollection is the same inbox problem with better handwriting.