Earlier this year Brad Shaver and I had a discussion about salinity and extractants.

@asianturfgrass Yeah, I agree with you about not using for fertility recs. But still confused about high soluble salts, K, and extractants.

— Brad Shaver (@MyTurfResearch) April 15, 2015

I had written previously this post explaining that a saturated paste extract is not a good way to look at soil nutrients and that it is not a good idea to look at a saturated paste extract and compare it to a standard soil test.

Brad asked about potassium (K) in saline soils, about acid extracts overestimating exchangeable K in saline soils, and alluded to a continuing confusion about the combination of high soluble salts in soil, potassium, and different extraction methods.

I’ll explain this in two ways. First briefly, without all the details.

Saturated paste (I’ll abbreviate as SPE for saturated paste extraction) is not useful to evaluate K in soils with high soluble salts because the problem with saline soils is too many soluble salts. The solution to this is leaching of the salts. The K measured by the SPE will be deliberately leached, and depending on how saline the soil is, a large portion of the K measured by a standard soil test, because it measures soluble and exchangeable K, will be deliberately leached as well.

Because one is going to deliberately leach soluble salts from a saline soil, as part of the standard management of saline soils, it doesn’t make sense to use the soil test K, from any extraction method, to determine how much K to apply as fertilizer to saline soils.

What does make sense? There will be some K in the soil. There will be some K added through irrigation water. And in a saline situation one can supply K as fertilizer in the quantity that the grass can use, disregarding the soil K and the K added in irrigation water. This guarantees the grass will be supplied with more than enough K, and one doesn’t need to test the soil for K at all.

Now explained the second way, with a few more details, and some data.

The purpose of soil testing is to determine if an element is required as fertilizer, and how much of that element should be applied. Or, in the case of salt-affected soils, the purpose of testing is to identify the problem and to determine what actions should be taken to solve the problem. Of course, if there is a problem with soluble salts, and one leaches them, it doesn't make sense to try to make a fertilizer recommendation from something one is going to be removing from the soil.

There are two forms of plant-available K in the soil: soluble and exchangeable. A SPE measures the soluble K and a small amount of exchangeable K. A standard soil test, such as the Mehlich 3 or normal ammonium acetate extractions, measures the soluble K plus the exchangeable K. Both the SPE and the standard soil test measure the soluble K, and the standard test will additionally measure exchangeable K.

Here are data from nine sites with the electrical conductivity of the saturated paste extract (EC_e) labeled as (ec), the K in ppm by SPE labeled as (kh2o), the K in ppm by Mehlich 3 labeled as (km3), and the location of the sample.

I've marked the EC_e = 4 dS/m level with a red line, to show in which cases a soil would be considered saline, and in which it would not. Note that one will try to maintain a site-specific EC_e depending on the species being grown and the irrigation water salinity -- the 4 dS/m level is included here as a reference level. These samples represent a range of soil salinity levels, most not saline, and two of them saline. Let's look at what happens with soil K across this range of soils and salinities.

The K extracted by SPE, which I have labeled as K_H2O to indicate it was extracted by water, is low when the EC_e is low, and it increases when the EC_e is higher. That is to be expected, because the quantity of soluble K is expected to be a function of the soluble salt content of the soil.

Now we can look at the Mehlich 3 K (K_M3) for these same samples.

This looks a bit different, as it should, because the Mehlich 3 test is measuring both the soluble K and the exchangeable K. When the soil salt content (the EC_e) is low, then the K_M3 is going to be influenced by the cation exchange capacity of the soil and the quantity of K on the exchange sites, and when the EC_e is high then there will be a greater proportion of soluble K as part of the the K measured by Mehlich 3.

This next chart demonstrates that. In each of these samples, the K_M3 is a larger value than the K_H2O. That is because the Mehlich 3 test measures soluble and exchangeable K, while the SPE test measures only the soluble K. By looking at the difference between the K_M3 and the K_H2O, we can see that the more salt there is in the soil, the smaller the difference is between these two quantities.

Is this making sense? When salt in the soil is low, which is what we want, there tends to be a big difference between the quantities of K extracted. As the salt in the soil increases, the difference gets small, because the quantity of soluble K is very high compared to the amount on exchange sites -- at least in a sandy rootzone.

Slight tangent for a moment -- this is something I've written and talked about before, as something that one should not be bamboozled by.

One wants to have low soluble salt content in the soil. When there is low soluble salt content, it is normal to have a large difference between the water soluble and the exchangeable nutrients. But that doesn't mean the grass won't be supplied with enough nutrients. From Environmental Chemistry of Soils (McBride, 1994): "Ion exchange reactions at surface sites exposed to solution are extremely fast."

Back to the data, now looking not at the difference between K_H2O and K_M3, but the ratio between them. Remember, K_M3 in these data is always larger than K_H2O, because K_M3 contains both the water soluble (K_H2O) and the exchangeable K.

With this proportion, when it is close to 0 (on the y-axis), that means the K_H2O by saturated paste is only a small amount of the K_M3. At low soil salinity, that's just what we see. And with increasing EC_e, as expected, the proportion of soluble K increases.

This can also be represented in a linear relationship for these data by showing that same proportion of $(K_{H2O}) / (K_{M3})$ across the natural logarithm of EC_e.

From this chart, it seems that knowing EC_e and K_M3 is enough to predict K_H2O. Not only is the K_H2O value not useful in making a fertilizer prediction because one will try to leach it away with the other soluble salts in a saline situation, but it can be predicted from other measurements, meaning it isn't adding any new information.