Two laboratories can receive the same soil sample and return phosphorus recommendations that differ by more than 50 kilograms of P₂O₅ per hectare. Neither is wrong. Soil test interpretation depends on three decisions layered on top of the measurement: the extractant, the calibration philosophy, and the yield assumption behind the maintenance rate. Understanding which of the three is driving a disagreement is what turns a confusing report into a decision.
Why do two labs give different fertilizer recommendations from the same soil sample?
Because a report contains two different things, and only one of them is a measurement. The number is the measurement. The recommendation is a policy built on top of it.
Most advisers blame the extractant first. That instinct is usually wrong. Of the three sources of disagreement, extractant chemistry is normally the smallest.
The clearest evidence comes from Zhang and colleagues, who compiled soil-test-based phosphorus and potassium recommendations from the 13 states of the SERA-IEG-6 group in a 2021 paper in the Soil Science Society of America Journal. Among the states using Mehlich-3 — the same extractant, the same units, the same crops — critical phosphorus concentrations for maize and warm-season grass hay ranged from 30 to 75 mg/kg. The chemistry was held constant and the threshold still varied by a factor of 2.5.
Mehlich-3 vs Bray P1: what the extractant actually changes
Mehlich-3 extracts more phosphorus than Bray P1 on most acid to neutral soils, and the relationship between them is strong but regional. Culman and colleagues at Ohio State analysed 2,659 samples from Ohio and Indiana for their 2020 paper in Communications in Soil Science and Plant Analysis. Mehlich-3 phosphorus averaged 35 percent higher than Bray P1, with the two tests predicting each other closely (R² of about 0.95).
That ratio does not travel. The same paper's literature review found published Mehlich-3 to Bray P1 ratios of 1.19 in Delaware soils, 1.4 in Missouri, and 1.53 in the Virginia Piedmont. A conversion built on glacial till soils in Ohio is a different equation from one built on Coastal Plain sands.
There is a practical complication most advisers have not registered. As Bergschneider and Margenot noted in a June 2026 farmdoc daily article from the University of Illinois, a Bray P1 value on a modern report was most likely converted from a Mehlich-3 analysis rather than measured by wet chemistry. The extractant printed on the page may not be the extractant that touched the soil.
Two situations do make the extractant decisive. On calcareous soils, the acid in Bray P1 is neutralised by carbonates and the test underestimates available phosphorus; Iowa State confirmed this again in its February 2023 revision of PM 1688 and directs users to Olsen or Mehlich-3 instead. And Mehlich-3 measured by ICP reads higher than Mehlich-3 measured colorimetrically, which is why PM 1688 publishes separate interpretation tables for the two.
Sufficiency vs build and maintain: the larger source of spread
Above the critical level, systems stop agreeing about what a soil test means at all. This is where most of the rate difference is created.
Iowa State's approach, updated by Antonio Mallarino in February 2023, treats interpretation categories as probability statements: roughly 80, 60, 25, and 5 percent chances of a yield response in the very low, low, optimum, and high categories. In the optimum category the recommendation is crop removal, calculated from prevailing yields — explicitly not from a yield goal. Above optimum, only starter rates in specific conditions.
The 2020 Tri-State Fertilizer Recommendations for Indiana, Michigan, and Ohio (Bulletin 974, by Culman, Fulford, Camberato, and Steinke) kept a build-and-maintain framework but narrowed it. Drawing on 198 farmer-coordinated trials across 39 Ohio counties from 2006 to 2018, the revision set the phosphorus critical level at 20 mg/kg Mehlich-3 for a maize-soybean rotation and 30 mg/kg where wheat or alfalfa is included. It also made build-up optional rather than expected, and removed the drawdown range, so recommended rates now reach zero at the top of the maintenance range.
The Illinois Agronomy Handbook goes further toward building, targeting Bray P1 levels of 40, 45, and 50 pounds per acre (20, 22, and 25 mg/kg) depending on the estimated subsoil phosphorus supply of the region.
Place all three on one soil testing 18 mg/kg Bray P1 under maize and soybean. Iowa calls it optimum and recommends removal. Tri-State, converting to roughly 24 mg/kg Mehlich-3, places it above critical and inside maintenance. Illinois, in a low-supply region, adds a build increment on top of removal. The measurement never changed.
A worked example: two Cerrado soils, one number
Brazil makes the point in the sharpest form, because tropical soils vary enormously in how strongly they hold phosphorus.
Embrapa Cerrados published the calibration in Circular Técnica 33 (Sousa, Nunes, Rein, and Santos Júnior, 2016). Using Mehlich-1 and targeting 80 percent of potential yield, the phosphorus critical level is 18 mg/dm³ on soils with 15 percent clay or less, and 4 mg/dm³ on soils above 60 percent clay. The phosphorus buffer capacity moves in the opposite direction: about 6 kg P₂O₅/ha to raise the test by 1 mg/dm³ on a sandy soil, and roughly 54 kg P₂O₅/ha on a soil with 61 to 65 percent clay.
Now take two fields that both report 8 mg/dm³ Mehlich-1, both going into soybean at 3 t/ha.
- Field A, 18 percent clay. Critical level 18 mg/dm³, so the soil is below it. Corrective rate = (18 − 8) × 6 = 60 kg P₂O₅/ha, plus maintenance at Embrapa's removal figure of 15 kg P₂O₅ per tonne of soybean, or 45 kg/ha. Total: 105 kg P₂O₅/ha (about 94 lb/acre).
- Field B, 62 percent clay. Critical level 5 mg/dm³, so the soil sits in the high class. No corrective application, and Embrapa reduces maintenance to draw on stored phosphorus: 30 kg P₂O₅/ha (about 27 lb/acre).
Same number, same country, same extractant, same laboratory. A 75 kg/ha gap, created entirely by clay content. At a local price of US$1.00 per kg of P₂O₅ that is US$75 per hectare; substitute your own price to size it.
The choice of extractant can make the correction unnecessary. Embrapa's data show that ion-exchange resin values are little affected by clay content, so a single critical range of 15 to 20 mg/dm³ applies across textures — the basis of the resin method used in São Paulo's Boletim 100. The CQFS-RS/SC manual for Rio Grande do Sul and Santa Catarina takes the opposite route, keeping Mehlich-1 as the reference and converting Mehlich-3 results back to Mehlich-1 equivalents through an equation containing clay percentage.
Can you convert between soil test methods?
Within the soil population the equation was built on, yes. Outside it, treat conversions as approximations rather than answers.
The same 2020 Culman review illustrates the limit. Across nine published studies, the Mehlich-3 value equivalent to 100 mg/kg of ammonium-acetate potassium averaged 107 mg/kg but ranged from 66 to 159 mg/kg. A single conversion factor would move a field across two interpretation categories depending on which study supplied it.
The calibration itself carries similar uncertainty. Slaton and 17 co-authors, publishing in the Soil Science Society of America Journal in 2024, fitted five model-and-interpretation combinations to one Olsen phosphorus dataset. The resulting critical values ranged from 7 to 16 mg/kg. One dataset, one extractant, one crop — and a critical phosphorus level that more than doubled depending on the curve chosen.
How to read a report generated under an unfamiliar system
Ask five questions before you accept the rate.
Which extractant, and measured how? Mehlich-3 by ICP and Mehlich-3 by colour are not interchangeable. Confirm whether a reported Bray P1 was analysed or converted.
What is the sampling depth? Illinois calibrated on 18 cm (7 inches); Brazilian systems generally use 0 to 20 cm. Depth changes the number before anything else does.
Which philosophy applies above the critical level? Removal, removal plus build, or zero. This decides most of the rate.
Where does the yield figure come from? Prevailing yield and yield goal produce different maintenance rates from identical soil chemistry.
Does the critical level adjust for soil texture or buffering? Essential in high-clay tropical soils, largely absent from temperate systems.
For readers working in the United States, the Fertilizer Recommendation Support Tool offers a way to check. Lyons and colleagues described its database in Agricultural & Environmental Letters in 2021, assembled from more than 1,200 correlation and calibration trials; the public tool launched nationally in 2024 and lets an adviser see the underlying trials for a given crop, region, and extractant.
The number on the report is not the recommendation. The calibration behind it is.
At Valora Earth, this is why our advisory service asks which extractant and depth a report used before returning a rate, in whichever of 40-plus languages the farmer is writing in.
Frequently asked questions
What is a phosphorus critical level? It is the soil test value above which the probability of a yield response to phosphorus fertiliser becomes low. It is a statistical threshold from field trials, not a fixed property of the soil. Critical levels change with extractant, sampling depth, crop, soil texture, and the model used to fit the response curve.
Is Mehlich-3 better than Bray P1? Mehlich-3 is more versatile: it works across a wider pH range and extracts several nutrients in one step. It is not more accurate on soils where Bray P1 is well calibrated. On calcareous soils Bray P1 fails and Mehlich-3 or Olsen should be used instead.
Why do Brazilian critical phosphorus levels look so low? Because Mehlich-1 critical levels fall as clay content rises. Clay-rich Oxisols bind phosphorus strongly, so a small extracted value can still support the crop. Embrapa's Cerrado calibration ranges from 18 mg/dm³ on sandy soils down to 4 mg/dm³ on soils above 60 percent clay.
Use Valora On-the-Go!
