Why Labs Have Different Soil Test Results and Recommendations

With the advent of grid sampling, many people are taking more soil samples, and paying more attention to them. Because of this, there are more questions about why soil labs get different results for the same field, and even the same sample, and why fertilizer recommendations can be significantly different between laboratories. There is a long, but reasonably simple explanation for these differences.

A soil sample report is really two distinct services:

  • The analytical results which are the amount of nutrients found in the sample.
  • The recommendation which is the amount of fertilizer recommended to be applied based on the crops to be planted.

Lab differences occur in both areas, so we must look at each area separately.

Analytical Results

The primary reason for significantly different analytical results is the different chemical extractants used by labs. At first, it may not seem logical for labs to use different chemicals to extract the same nutrients, but there are some good reasons for this. An example of this is phosphorus. The following list, which is in rough order of method introduction, contains most of the ways that phosphorus may be extracted.

  1. Bray & Kurtz P1 (The original “standard” extractant, developed for acid Midwest soils)
  2. Bray & Kurtz P2 (A stronger version of P1 that identifies less soluble P, due to rock-phosphate use)
  3. Olsen (Developed for high pH soils, where the Bray & Kurtz methods were thought to be weak)
  4. Morgan (Developed in the Northeast States as a more “universal” extractant for acid soil)
  5. Modified Morgan (An improvement, to include micronutrient analysis)
  6. Mehlich 1 (Developed for the acid, low CEC Southeast soils)
  7. Mehlich 3 (A modification of Mehlich 1 for higher CEC, Midwest soils)
  8. Resin (Under development in Brazil for tropical soils)

Each of the above methods for extracting P from soil was developed for a good reason, and is an excellent “predictor” of P availability when used appropriately. However, they will typically give considerably different numerical values for the soil P level. A similar example could be given for most, or all of the other nutrients in a soil test. A number of agronomists have developed mathematical formulas for converting the results of some of these methods to some of the others. While there is normally a relationship between the results obtained by the different methods, it may not be possible to accurately convert the results of one method to another. It is possible that there will be multiple correlation's between any two methods for different soils, or soil conditions (such as pH).

Some other, typically less significant, reasons for differences are

  1. The type of instruments used to detect the nutrient in the extractant.
  2. The quality and techniques used by the lab employees
  3. The method of reporting the results.

A common problem in understanding soil test reports is in understanding the relationship between the different units of measure that labs use in reporting results (item 3, above). The following is a list of the common reporting units and the factors that can be used to convert them.

  • Parts per million (ppm) × 2 = Pounds per acre (lb./ac)
  • Lb./ac of P × 2.291 = lb./ac of P2O5 or lb./ac of P2O5 × 0.4364 = lb./ac P
  • Lb./ac of K × 1.2046 = lb./ac of K2O or lb./ac of K20 × 0.8302 = lb./ac of K

Another common practice is to report results as an “index”. Labs sometimes do this in an effort to improve the relationship of the soil test to the availability of the nutrient to the plant. Most people in agriculture understand that there are many factors that affect the availability of nutrients to the crop. These factors are in addition to the amount that is measured by the soil test. When the agronomists at a lab believe that they have discovered a way to include some of these other factors, they develop formulas that modify the analytical result. These modified results appear on the soil test report as an index, and are normally labeled as such. There is nothing wrong with using an index system, if it improves the relationship between the soil test and the crop response. Since an index is unique to the lab that developed it, it is often difficult to correlate with another lab's results. With all of these complications, the analytical side of a soil testing service is still typically more accurate and precise than recommendations. With nutrient management planning there are also some states that are indexing the phosphorus number. Most of these index numbers have been developed as a compromise between environmental agencies which deal with water quality issues, NRCS specialists and state land grant university agronomists.


The agronomy that serves as the basis of making fertilizer recommendations is, by nature less accurate than the chemistry underlying the analytical service. To understand the problems inherent in making recommendations, we should review how agronomists approach the task.

The first consideration is to decide on a recommendation philosophy. The two most common choices are

  • Minimum fertilizer needs (sometimes called “crop removal”)
  • Soil build-up

The choice between the two philosophies is not as simple as it might seem, because soil science does not completely understand every aspect of soil chemistry, and many unknown or unmeasured factors will have a large affect on the performance of the recommendations. Some of these factors include rainfall, average temperatures (growing degree days), varieties, plant population, rotation, weeds, pests, diseases, depth of topsoil, and many other factors.

Minimum Fertilizer Needs

This is the short term philosophy of fertilizer recommendations. Reasonably good estimates of the amount of each nutrient that crops remove from the soil have been developed. In its simplest form, this approach recommends the amount of nutrients that will be removed for a given yield goal. A refinement of this approach might include some factors to account for the losses and inefficiencies that occur between the fertilizer application and the plant uptake of those nutrients. Doing this means that we are depending on the native soil fertility to supply the nutrients to grow the roots, stems, leaves, and other plant parts that are not harvested. These crop parts often contain the majority of certain nutrients. If we don't do a good job of producing the un-harvested plant parts, we won't get the harvested portion that we are after. Therefore, if the native soil fertility does not produce adequate stems, leaves, and roots, we are not likely to achieve the desired yield of grain or fruit. Another consideration is that we are applying the fertilizer to soil, not injecting it directly into the plants. This means that the soil and fertilizer will interact, and much of the applied nutrients will not get into the crop that season. We cannot accurately predict the amount of fertilizer that will be tied up by the soil. If this tie-up is more severe than we predict, the crop will be starved for one or more nutrients, and yield will suffer. Applying “just enough” fertilizer can be like trying to take a long drive with “just enough” gasoline to get there. It will work if we are smart enough to know exactly how much gas we need, all of the problems that will happen, and the exact effect that each problem will have on our mileage. Fertilizing a crop this way typically costs less initially, but can be a high risk approach, if the underlying soil test is not excellent.

Soil Buildup

This is the longer term philosophy of fertilizer recommendations. It is probably best suited to land that is owned by the farmer, and should be evaluated by the average yield and profits over several years. The logic and science behind this approach is that crops will grow better and yield better when all of the roots are exposed to an abundant supply of nutrients over the entire life of the crop. The only way to accomplish this “ideal” situation is to have excellent soil test levels. With this approach, much of the recommended fertilizer in the early years of the build-up period may be in excess of the crops needs, and will go toward increasing the soil test levels. Using this approach should produce higher yields in the early build-up years, but perhaps not in proportion to the added costs. However, as the soil nutrient levels increase, yields and profits should increase faster than costs. Of course, for this to happen the farmer must manage the crop properly to insure that other factors do not hurt yields. Weather will always affect yields, but crops on good soil tests will typically suffer less in bad years, and take better advantage of good ones.

When agronomists talk about increasing soil test levels, they typically mean the phosphorus (P) and potassium (K) levels. However, it is not difficult to increase soil test levels of zinc (Zn) and copper (Cu) also. The level of sulfate sulfur (SO4-S) can be increased somewhat, although the bigger effect might be on the sub-soil, due to the moderate mobility of SO4-S. Nitrogen (N) and boron (B) are mobile enough that we don't try to build up soil levels, and iron (Fe) and (Mn) availability is so strongly affected by soil pH and other factors that building the soil test levels is not normally effective.

Increasing the soil levels of P and K involves overcoming the soils ability to “fix” much of the applied nutrients into unavailable forms. Each soil has a different ability to fix P and K, so the amount of fertilizer required to overcome this will be different for every soil. A rough rule-of-thumb used by agronomists is that it will take 9 or 10 lb./acre of P2O5, and 3 to 4 lb./acre of K2O to increase the soil test level 1 lb./acre of P or K respectively. Remember that these nutrients are in addition to those removed by the crop. The soil test level at the beginning of the buildup process may also affect the rate of soil test buildup. Research and experience indicate that a soil with very low levels of P or K will require several times as much fertilizer nutrient per unit of soil test increase, compared to a beginning soil test level that is higher. If you do the arithmetic for a buildup program, you will see that it results in some very high fertilizer recommendations. Most agronomists spread the buildup over 3 to 5 years to make it more economically feasible. If you begin to review the recommendations of most labs, you will find that they don't always use these exact buildup figures. This is because other valid data may be factored into the recommendations.


After deciding on the total amount of fertilizer to recommend, most labs will begin to make adjustments for a few other factors such as crop rotations, manure or sludge applications, and other factors. While there is rough agreement on the general effect that these factors have, there is probably no agreement on the exact values to add or subtract for each factor.

Many labs will also make adjustments for fertilizer application methods. Application methods can have a large effect on the total amount of fertilizer recommended. Many farmers apply part of the fertilizer in concentrated bands that greatly increase the uptake efficiency of the nutrients in the band. Other factors that can change recommendations include irrigation, nutrient injection into irrigation water (fertigation) and the use of foliar applied nutrients in low volume sprays. Each of these application systems can change the efficiency of nutrient uptake by the crop, and therefore the amount of fertilizer needed.


In the everyday world of making recommendations for farmers, you will find that you are often required to make adjustments to laboratory recommendations because you have unique information that the lab cannot know. This is the way it should be, because no laboratory service can be expected to have all of the answers, or to make the best recommendations for all situations. One consideration that can be a basis for local adjustments is the financial and management conditions of the farmer. Many farmers simply cannot afford a strong buildup program, or they are often farming land that is rented on a year-to-year basis. In these cases, some intermediate recommendation may be the best. Soil testing labs do not typically consider these factors in making recommendations, but somebody probably should. Unfortunately there is no formula that the lab or you can use to make a “perfect” recommendation for each situation. This is where the knowledge and experience of the local person working with the farmer becomes the difference between successful recommendations and those that are not.

library/articles/why_labs_have_different_soil_test_results.txt · Last modified: 2010/03/30 16:40 by bill