The Fallacy of Mainstream Potassium and Nitrogen Fertilization with Richard Mulvaney

Hi, friends. Welcome back to the Regenerative Agriculture podcast. Today I'm here with a soil scientist and researcher who I've heard quite a bit about, Richard Mulvaney from University of Illinois in Urbana. Richard has done some interesting work looking at nitrogen and potassium in soil profiles and the value that we get out of fertilization and also trying to better understand those environments where fertilization may be less necessary or not produce the crop response that we anticipate. Richard, I'd love to hear more of your story and your background. Can you tell us how you came to this space and the journey that brought you to looking at fertilization and the impact that it may or may not have on some soil and plant ecosystems?

Sure thing, John. My journey took many years and in fact it's still ongoing. But I began working in this profession back in the 80s and actually 70s as a graduate student. And at that time I was totally focused on learning more about the processes that lead to the loss of fertilizer nitrogen and, and reduce its effectiveness for increasing yield. That's where my focus was. I did my graduate work related to that view and at that point I got involved in using isotopic nitrogen as a tracer. That's N15. But again, the emphasis was always on minimizing the loss of fertilizer N to increase its availability for crop uptake. It was always from that perspective. And in fact it's not surprising because that's the perspective that was so prevalent in the literature. The concern was always about how do we reduce the loss of the fertilizer. Beginning in the 1990s, I, I began to come to a different view, collaborating with my colleague Saeed Khan. We were working on nitrogen back at that time and we had some connection to a PhD project here in Illinois that was finding that there were quite a number of on farm sites in the state where there was no statistically significant yield response by corn to fertilizer nitrogen.

That's not supposed to happen.

It's not supposed to. I had been taught, and I hate to admit it, but I actually at one time had actually taught this in my classes, that the optimal rate of fertilizer nitrogen is easily calculated by multiplying 1.2 times an expected yield goal and then deducting nitrogen credits like manure or a previous legume. That was the standard way that nitrogen rates were recommended in this state and in fact in many others. It was a yield based approach and I had never given that approach a whole lot of thought. But I had to begin wondering about this, when those non responsive sites showed up. And in fact there were quite a few of those non responsive sites. Out of 75 site years that were studied in that PhD project, I believe there were 33 that showed no significant response to N. It's quite a proportion.

Just for some context, did these areas that were non responsive, did they have strong yields? What were the yields like in these areas?

Yes. And what would typically happen in those sites is that the unfertilized yield or the check yield would be very high and would not be increased significantly by any application of fertilizer nitrogen. That was the typical pattern.

That's enough to give some fertilizer executives heartburn, I imagine.

I imagine it was. And so that was kind of a wake up call for Said and myself. And it was giving us a different view that this 1.2 calculation may not be quite so reliable as we had assumed it was. In fact, it led to a total change in thinking on our part. But that took a while to happen. So by about the year 2000, we were engaged in a project to try and chemically characterize some of those non responsive sites as compared to responsive sites that had been studied in that PhD project. We had the soil samples and we were chemically analyzing them. And that brought the first revelation to my thinking that there was something far more wrong here than I had ever realized. And the problems began with the analytical methods used to characterize organic soil in. I had worked with two of the big names in that area who developed those methods, and it came as a bit of a. A shock to find out that they didn't work. And they didn't work. Yes. So the methods that we had based our chemical characterizations on turned out not to be reliable or quantitative. And Saeed and I went to work reinventing those methods with a very different approach. And that took us about a year to do. Let me preface that progress by telling you that when we had started that work to look at the comparison of responsive and non responsive soils collected in that PhD study, we had found no differences. That the responsive soils looked just like the non responsive soils. That in fact was very consistent with what had been seen years before. When those fractionation methods had been applied to other soils and other studies, everything looked about the same. So I talked with my wife at that point and she made the very interesting comment that, well, now isn't that interesting. You saw scientists are saying that the soil organic matter is made up of course, carbon and nitrogen, and that the carbon coming from plant materials, it's all Different. It's very heterogeneous. So you have simple sugars and proteins and cellulose and hemicellulose and lignin and all that. And those constituents are very different in their rate of decomposition by the microbes. They're different. For example, the simple sugars and proteins go very quickly and the lignin is the last. They're all different. But she told me now when it comes to nitrogen, you guys are saying it's all the same. She was right. That's what the textbooks had said. There's no fraction that is more decomposable than any other in terms of N. And that's what we were finding with the beginning of our work. To compare those responsive and non responsive sites, then you need to realize that the plants had seen a difference. The plants at the non responsive sites did not need fertilizer. They got all they needed from the soil. I'm talking about nitrogen. And the ones at the responsive sites did need fertilizer. The soils were different. There must be a chemical difference. But the methods we were using did not detect it. They were flawed. So as I said, Saeed and I set out to work to develop methods that would replace those and make it possible for us to characterize that organic soil in. We did that and it took about a year as I said. Then we began to apply the new methods that were based on a process called diffusion. We applied those methods to again some of those same soil samples, comparing responsive and non responsive sites. And lo and behold, a pattern emerged. I never expected to see it. I believed my former professors, what they had written the textbooks and so forth, said there's no difference. But there was a difference. And I can still recall said coming in one Monday morning and saying I found a difference. Out of the fractions that we had analyzed, there was only one that showed a clear separation of responsive and non responsive soils. The non responsive soils were all testing higher than the responsive soils. And that fraction is one that is typically referred to as amino sugar nitrogen. That was a radical change in thinking for me. I realized that something else I had been taught over the years, this is a great fallacy. The fallacy is that the fertilizer is the main source of nitrogen for crop uptake. I'm talking especially about corn. And the soil with the process of mineralization is only minor, that the soil supplies some nitrogen, but it's pretty minor in terms of the quantity. And the fertilizer is the dominant source. That idea was totally consistent with that 1.2 method of n. Recommendation which emphasized the importance of the fertilizer, same idea. And now, lo and behold, we discovered that soils differ. Some of them have a lot more of that amino sugar fraction and presumably can supply more nitrogen for the crop, even to the point where they no longer need any fertilizer.

Wow.

That was a total transformation of my thinking from that time on. I was never able to go back to the 1.2 idea. I could not believe in that anymore.

Well, there are many farmers and many agronomists who have observed and have experienced infield experiences where as you identified, there is no correlation between the measured soil nitrogen as measured by typical historical lab analysis. There's no correlation between those measured nitrogen levels and actual crop response to applied nitrogen.

No.

So I think that's becoming a more mainstream awareness. But I'd like to further unwrap what you described as the fallacy that fertilizer is the main source of nutrients, particularly nitrogen, in corn in this case. So if fertilizer is not the main source and amino sugars are the main source, or I should say perhaps in more broadly general soil biology and biologically derived nutrients, what proportion of the applied nitrogen is actually delivered to the crop?

Well, typically and again I'm going back to the isotope N15, which we've done quite a lot of work with and others have too, and over and over and over again that isotope will tell us. We weren't listening back in the earlier days, but the data are there and they're telling us that typically the soil which would be the unlabeled end that ends up in the crop that would typically supply. Again, it varies, but typically on the order of at least 2/3 of the N in that crop at harvest, at least 2/3. And with one of those non responsive soils I was mentioning, it could be close to 100%.

So on these non responsive soils, applying nitrogen fertilizer is a complete total waste of money.

Yes, that's what it means, that the crop didn't need it and the farmer should not have applied.

That's really enough to upset the fruit basket in a really significant way for mainstream agronomy.

Well, we found that was the case too. And so it gets. There's one more aspect to that 1.2 method I need to mention because this is something we didn't know back in 2000 and in fact it took several more years for us to begin to understand this. But it so happens that the 1.2 method comes from, at least in Illinois, it comes from university research trials that had been conducted in the late 60s. And those trials were directed toward optimizing economically the end fertilizer rate for corn. But a really important point about this is that those trials had been designed, as I said, on university research fields as static plot studies. That's very important. And I now emphasize that in my fertility class, the trials were laid out and the data were collected in multiple years at the same sites. And when I say static plots, I mean that a given plot area in that study received the same fertilizer treatment year after year. That's especially important for the unfertilized treatment. The unfertilized treatment gets no input of fertilizer and it grows corn that removes the nutrients every year. There's no replacement. The soil is being depleted, being mined. But more importantly for nitrogen is the fact that the carbon inputs coming from the residues, those are going to be broken down by the microbes. But in so doing, the microbes need and must have nitrogen has to. They're going to take it up from the soil, they're going to deprive the next crop of nitrogen and they're going to depress the yield. That's what's going to happen. And so in a static plot study, you accentuate the value of the fertilizer by depleting the soil. I hope that makes sense.

That is very clear. And it's in alignment again with the experiences of many agronomists and growers we've had conversations with who have totally debunked the 1.2 number on their own operations, where there are now many growers who are growing Rudic grade crops, in some cases with no applied nitrogen, and in some cases with applied nitrogen on the order of 0.4 to 5 pounds per bushel of nitrogen credits.

Absolutely. And so let me make one more point about that static plot design. The 1.2 method had been developed on the basis of that design. It made the fertilizer look better than it should have. Then that 1.2 method was applied to farmer fields. The farmer fields were not managed or fertilized according to static plots. There was never an unfertilized area that was left year after year.

Never.

And so in that PhD study I mentioned where they had found these non responsive sites on farmer fields. Lo and behold, we came to realize, well, that's in large part due to the fact they developed the 1.2 method on static plots and now they take it to the real world of farm fields. And the soil supplies much more in than the method recommended. It's exactly what you'd Expect.

Yeah. This finding is obviously quite remarkable. What you're describing is that 33 out of 75 locations are non responsive. 100% of the nitrogen is being supplied from the soil, not from applied fertilizer. And in the remaining 40 some plots, approximately 66% of the nitrogen in the crop at harvest is still coming from soil. Which means only 33% is coming from applied fertilizer.

Yes. And so let me make one more point about the 1.2 method. And that was it had an underlying assumption that was not obvious and we didn't realize it for many years, but finally dug it out. And the assumption behind the whole thing was that one third of the nitrogen was assumed to come from the soil and 2/3 from the fertilizer. It's upside down.

Understanding that it's upside down, what is the real world scenario and how do we optimize what actually happens in the real world? If it's the case that 66% to 100% of the nitrogen within a crop can come from organically derived nitrogen that is in the soil, perhaps amino sugars. I personally would love to learn more about amino sugars. And then how can we optimize that? How can we develop more amino sugars in our soil profiles?

As we gained an understanding that the soil really is the major source of nitrogen for crop uptake, it just stands to reason that you should test the soil, if that's the main source, in order to determine how much fertilizer you might need to apply. And that's where we went. So we developed the Illinois soil nitrogen test. That was an estimate of this so called amino sugar fraction, and then that was oriented toward variable rate nitrogen. Soils differ not only between farms, but even within the same field. There were entrepreneurs in the private sector that began to use this kind of a test and still do use it for variable rate and recommendations so that a soil testing higher in this fraction gets a lower fertilizer rate and vice versa. That's where it went and that's where it still needs to go.

Where is this soil assay? The Illinois soil nitrogen Test. Where is this available? What laboratories make it available to growers at this point?

One of the labs that runs it is actually headed by a former student of mine and we're doing on farm research in conjunction with them on several projects this year. But they routinely run this Illinois end test to do variable rate. N and the firm's name is Cropsmith. Kim Smith is the lead agronomist there. Cropsmith is the firm.

Awesome. Thank you for Sharing that. So what exactly are amino sugars? What proportion of the total nitrogen that is delivered from the soil is coming from amino sugars? What are amino sugars and how do we develop more of them?

Okay. Amino sugars, like other organic forms of nitrogen, are products of microbial activity. They're biologically synthesized, and they occur in various microbial constituents, like microbial cell walls. They occur in spores, and they also occur in the material called chitin. We think of insect exoskeletons or chemicals are made of chitin. Now, I'm not saying that the chitin is easily broken down, but I do believe that those other forms of amino sugars, namely bacterial cell walls, I think those are quite a bit more decomposable. And there's another fraction that contributes to this amino sugar n, which actually isn't amino sugar, but it occurs in the form of two essential amino acids. Those amino acids are asparagine and glutamine, and they contain two nitrogen atoms, not one, not just in the amino group, but in an amide group on the end of the chain. And that amide group is prone to break down. And it would be estimated by this Illinois soil end test, as well as things like glucosamine or galactosamine that occur in those bacterial cell walls. So that's a partial answer to what you raised the question of how much of the organic nitrogen in soils occurs in this form. The textbooks tell us 5 to 10% of soil organic and is in the form of amino sugars. Considering that those estimates come from the classic methods that we found were not accurate, I'm inclined to think that the proportion is actually higher than that. We've never done our own study to fractionate it and compare the fractions, but I do wonder if that number is a bit on the low side.

Probably reasonable to expect that the fraction of total nitrogen that is represented by these amino sugars will also vary based on the degree of soil biological health and activity.

Exactly. And so this amino sugar fraction has gained some traction as a parameter related to what is a term I don't like all that much. But here it is soil health. And I'm thinking of Solvita and Will Brinton, and we've been in contact with them, we've done some work with them. He has a Sland test that is another form of the Illinois soil end test, and it's targeting this amino sugar fraction. And so I totally agree with your statement. Yes, the more biological activity you have in the soil, the you'd expect this fraction to be increased.

I find your comment particularly intriguing about the source of amino sugars and your comments about a significant portion of these sugars possibly coming from cell membranes. I've been delighted recently to discover Dr. James White's research from Rutgers University where he describes what is termed as the rhythophagy process, where growing root tips are absorbing or internalizing entire microbial cells in large quantities and are internally within the growing root tips. They are exposing these bacteria and other organisms as well, but particularly bacteria, to superoxide and stripping the cell membranes. Then some of those bacteria, the entire bacterial cells, are consumed and a few of them survive the process and are then transmitt through the root hairs back into the soil environment along with root exudates that stimulate the reformation of their cell membrane. So the rhizophagy process is essentially a description of plant roots stripping off the cell membranes of all these organisms which you are describing gives them a source of amino sugars.

Yes. So again, to clarify, the amino sugar fraction that I'm aware of in terms of its location is in the cell, but again, that would be the outer external part of the bacteria and yes, it would be stripped off. Yes.

Fascinating connection. Now we come to the important part. For us as farmers and for those of us who might have soils that don't have soils in Illinois that don't offer nitrogen response, we have many soils in various parts of the world that do offer significant nitrogen response. How can we build the profiles of amino sugars in our soil profile? There's a growing awareness of the need for cover crops and to have constantly green growing plants and to incorporate manure and livestock and so forth. I think that's a solid foundation that many of our listeners have as a starting point. So let me perhaps refine the question a little bit to say, can we accelerate the development of amino sugars with applied nitrogen? Does applied nitrogen remain in the soil profile and contributes to an increase in the biologically available nitrogen?

Okay, when you say applied nitrogen, you're referring to synthetic fertilizer nitrogen?

That's correct.

Okay, yeah. That opens a door to another topic. And. But before we go there, let me make the comment that the increase in this so called amino sugar nitrogen fraction. I say so called because it's a mixture of different chemical forms, as I think I tried to explain. And a better term, and one that's more chemically descriptive, would be alkali hydrolyzable nitrogen. And that's how the Illinois soil N test estimates it. But the issue is how can that fraction be increased in the soil there. I'd make a couple comments. One is that in the early work, as we came to realize the importance of that fraction and its presence, we were finding that.