University of Minnesota Nutrient Management Podcast Episode “On-farm research and nitrogen rates”
January 2021
Written transcripts are generated using a combination of speech recognition software and human transcribers, and may contain errors. Please check the corresponding audio before referencing content in print.
(Music)
Paul McDivitt:
Welcome back to University of Minnesota Extensions Nutrient Management Podcast. I'm your host Paul McDivitt, communications specialists here at U of M Extension. Today on the podcast we're talking about on-farm research. We have three members of Extension's Nutrient Management team and a special guest, independent agronomist, Matt Wiebers. Can you each give us a quick introduction?
Dan Kaiser:
This is Daniel Kaiser. I'm an Extension Nutrient Management specialist with the University of Minnesota. I'm located out of the St. Paul campus and one area that I work on a lot are the nutrient guidelines for the majority of the agronomic crop species in Minnesota.
Fabian Fernandez:
I'm Fabian Fernandez also at the University of Minnesota based on St. Paul campus. Work on nutrient management for corn cropping systems focused on nitrogen management and environmental quality. And I contribute quite a bit of data into the guidelines that Dan mentioned.
Jeff Vetsch:
This is Jeff Vetsch. I'm a soil scientist and researcher at the University of Minnesota Southern Research and Outreach Center.
Matt Wiebers:
And I'm Matt Wiebers. I'm an independent agronomist here in east-central Minnesota, been an agronomist for about 25 years.
Paul McDivitt:
All right, Matt, can you start us off with a little background on your AFREC funded on-farm research project and what you've seen in the last few years?
Matt Wiebers:
Sure. So AFREC has funded this project since 2017, and continues to fund it here through 2021. Back in 2017, their proposal for RFP was looking for some on-farm research that looked at precision agriculture technologies and how those precision Ag technologies can add to the database of on-farm research. So what I did was pooled together a group of crop advisors and growers from around the state to put together a project. And we've done that in now 2018, 19, and just wrapping up the 2020 season.
Matt Wiebers:
The project mainly works around nutrients of nitrogen. We've expanded a little bit into sulfur and potassium over the last couple of years, but, with nitrogen being the nutrient of focus, that's predominantly where most of our on-farm research trials are. The project is looking at how we compare the MRTN rate of nitrogen, which is the guideline for how farmers should be recommending nitrogen rates for their corn crops in Minnesota against some rates that are above and below that. We're trying to answer the question of how good the MRTN is at predicting the right rate of nitrogen for farmers to use, to maximize yield and also economics.
Matt Wiebers:
So with this project, we're able to do a lot of research and fairly low cost. It certainly isn't free, but we're trying to come up with another way to do research besides what our friends at the university have been doing. The technology has really matured over the last two decades, and they're getting to be lots and lots of questions that farmers are needing to answer on their own farms. Shrinking budgets at the university are forcing farmers to do some of this work on their own farms.
Matt Wiebers:
And now with the nitrogen fertilizer rule being brought into play here, I think some farmers are really wanting to generate some data on their own farms in case they should need to use that in the future to defend or backup and support the choices that they're making on their farms. So, Paul, that's a little bit of an overview on on-farm research.
Paul McDivitt:
Yeah. And can you tell us about kind of some of the key findings that you saw? I think most of it was from that 2018 season, if I'm correct.
Matt Wiebers:
In 2018, we did the bulk of the research with 48 trials. 40 of those were nitrogen rate trials. And what we did was compare plus and minus 30 pounds to the MRTN rate. So that was our big year. And of course 2018 was a better growing season than 2019 when we had a lot of rainfall. But what we saw in 2018 was that the MRTN rate did actually work out to be the best rate in the bulk of cases. So if I look at the data here, we used $3.50 for corn and 35 cents for nitrogen. So that came up with that 10 to one ratio. And that MRTN rate was around 140 pounds of nitrogen for corn soybeans and slightly higher, I think, upwards of 150 to 180 for corn on corn trials.
Matt Wiebers:
So we did find that in that MRTN rate, we were close to maximizing yields. There was a little bit of an increase in yields as we added that 30 pounds, I would say it looks to be around three bushels per acre, so that just barely would offset the additional cost of nitrogen. If we look at cutting back 30 pounds of nitrogen, as we did in 2018, that had the tendency to be a little bit more of a steeper cliff. I think we decreased yields by closer to five bushels across all trials there with the corn on corn trials being a little bit more dramatic. I think we saw a reduction closer to 10 bushels per acre when we reduced the nitrogen below those MRTN rates.
Paul McDivitt:
And then for the corn on corn plots?
Matt Wiebers:
Right. The corn on corn, we didn't have as many of those. A lot of our research ended up being concentrated in Southern Minnesota and Southeastern Minnesota in particular. And I think that's because that's where farmers are used to doing on-farm research. It seems like there's a lot of pressure in that part of the state because of the car's topography and a lot of watershed initiatives for farmers to do research. So it was easier to put trials down in that area.
Matt Wiebers:
But I think that it does appear that corn on corn management, when it comes to nitrogen is a little more sensitive. I think farmers have to be certainly turning up their nitrogen rates from where they are with corn on soybeans, but it does seem like corn does respond to nitrogen, especially in those corn on corn situations. So I do see farmers, at least in the on-farm research we have, wanting to press towards the top end of those nitrogen rates. Certainly, I don't find very many farmers being conservative when it comes to nitrogen rates, especially in corn on corn.
Paul McDivitt:
And how did that kind of compare to the MRTN rates for corn on corn?
Matt Wiebers:
So the MRTN, if we look at 150 to 180, I think most of the growers were near the top end at 180. And some even tried to be a little bit above that, but for research within AFREC, we tried to coach the growers to select at least for the trial area, a rate that was within the MRTN zone so that we could test above and below the MRTN ranges there. And I think that most growers who manage their corn on corn ground carefully, they have to manage residue, disease and weed pressure differently than corn on soybeans.
Matt Wiebers:
I think those growers can actually achieve equal yield potential. At least our data would support that we can achieve good yields or great yields in corn on corn, as long as guys are managing that. So, that's a whole approach more than just nitrogen. I think it's fertility and a whole host of things. But I think we do see that high yields are potentially possible. And across the state from North to South, we see yield potential being the same, no matter which BMP region we're in. There's 210 to 220 bushels of yield potential on a plot average that we saw in 2018 across the whole state from North to South and from East to West.
Paul McDivitt:
Okay. Anything else you want to add about your study before we move on to the university side?
Matt Wiebers:
Well, I think if we look at the 2019 trials, everybody remembers that that was a little bit of a wet year. If we compare what we did in 2018 to 19, we changed the project a little bit to step it up and compare plus and minus 50 pounds of nitrogen in 2019, that was some coaching end advice that I had received. And so we were testing a range of 100 pounds of nitrogen in 2019. And so we did see a lot steeper cliff where guys have to be a little careful, you don't want to be peeling back your nitrogen rates too much when you're doing plus and minus studies by 50 pounds.
Matt Wiebers:
So if we look at the studies, there 12 of 17 or 70% of the trials did show a significant yield response to nitrogen 30% did not. And that's kind of one of the interesting things is, even when you're doing plus and minus 50 pound swings in nitrogen, almost a third of the trials still wouldn't show a statistically significant difference in yield. Contrast to our plus and minus 30 study the year before, we had a lot fewer studies that had statistically significant yield responses there, just because we weren't doing as wide of a swing in nitrogen rates.
Matt Wiebers:
Only a third of the trials in 2018 had a statistically significant yield increase and two thirds did not. So there is something interesting here with nitrogen, not necessarily being the only answer to high yielding corn. So I don't have all the answers, but that's really what we found with the on-farm research studies, Paul.
Paul McDivitt:
Yeah, really interesting. For the rest of the group, how does your research and results compare to Matt's?
Dan Kaiser:
Well, just kind of looking at some of our data, what Matt's seeing, particularly with corn on corn kind of echoes a lot of the stuff that the things that we've been seeing. And I know, I think Jeff could probably talk more about that because he's been doing awesome work in the Southeastern part of the state specifically. But if we go to I'm looking at like south-central Minnesota, and Southwest Minnesota, just looking at some of Fabian’s data. I mean, we can see some pretty substantial yield increases to nitrogen with corn on corn, which we expect.
Dan Kaiser:
And it's been one of the bigger questions in my mind is moving forward what those recommendations need to be, because with the MRTN system, the way it's built, it's all built around data. So essentially every year we put in data, which adjusts the rates based on the data going into the particular data set. So over time, particularly we see these years where we have increased nitrogen demand. What we should be seeing is a gradual increase in our MRTN values, which we have seen in corn on corn. Just in looking at some of our data, looking at the database since about 2000, I've seen what looks like a linear yield increase from the MRTN values and then we still continue to see this.
Dan Kaiser:
The thing about corn following soybeans, just having that crop, whether it's just the fact of not having the residue out there which can tie up nitrogen with corn residue or the fact that the soils are black and warmer, which may mineralize nitrogen quicker. They just seem to be more buffered with that. And we've seen some increase in need for corn following soybean, but overall, that one's been pretty close. So you're looking at our numbers. Just looking at widespread issues, it's been corn on corn, it's been kind of the main problem that we've seen in terms of increased nitrogen demand.
Fabian Fernandez:
Yeah and I was looking at some of that data again from the last few years. And as Dan mentioned, we kind of see very similar kinds of trends as Matt was mentioning from the studies that he's doing, where especially in corn on corn, which we are needing more nitrogen than in the past. And there are a number of things I think that are impacting that, because we don't see as much of that in the corn soybean.
Fabian Fernandez:
I suspect that there is quite a bit to do with crop residue. In continuous corn, we have a massive amount of corn residue that takes quite a bit of nitrogen to decompose and so, it starts creating immobilization of nitrogen and so that's where we see that need for additional nitrogen. And I think that has a lot to do with the fact that the new hybrids that we are using are a lot more resilient to the composition, they're protected. They come to the end of the growing season, pretty much intact. If you look at corn 15, 20 years ago, they will be starting to decay even before harvest because they had disease and damage and stuff like that, whereas now they're pretty much intact. And so it takes quite a while to decompose, and I think that, that's one of the major reasons.
Fabian Fernandez:
The other thing that is kind of interesting especially as we look at water quality research is the fact that nitrate leaching occurs in relationship to precipitation. The more precipitation you get, the more nitrogen that we are losing. An interesting thing that we have been noticing is that it used to be that most of the nitrogen loss in drainage, in tile drainage, was in the spring. Now we see that it's in the spring and again later in the fall. And so I think these two things combined, the immobilization that is taking place and potentially more nitrogen loss because of the wet conditions that we are experiencing are resulting in these increased nitrogen rate that is needed to optimize the yields.
Jeff Vetsch:
Well, I think Fabian touched on some of the... And Dan touched on things that we're seeing in South central Southwestern, Minnesota. I've been involved in a study that started in 2015 and it's grand through 2020 in Southeast Minnesota with our department of Ag partners. As Matt said, there's a lot of concern about nitrogen in groundwater in Southeastern Minnesota and the coast topography. So we're doing studies that are similar to what Matt did and they're on-farm, but we're looking at full N rate studies with three treatments that also look at split applications. So we're comparing not only with finding out what the optimum rate is, but also is there a benefit to changing the timing of application?
Jeff Vetsch:
So we've had 20 site years of data in this five-year period, 10 sites of corn following soybeans, and 10 of corn on corn. So when we look at the optimum calculated rates from those trials, when we look at the corn following soybean, we had one site that was below the acceptable range from the MRTN database, which is 118 to 144 pounds currently. We had three sites that fell within the range, three sites that were just slightly greater than the maximum.
Jeff Vetsch:
So between seven and 12 pounds greater than that 144. And we had three sites that were significantly greater than the maximum, between 29 and 59 pounds. As Fabian mentioned, two of those three sites that were significantly greater had excessive rainfall in spring months, that critical period for leaching and denitrification, which are may and June primarily. It was interesting as Matt said that there could be some significant yield penalties, these three sites that had economic optimums that were considerably greater than the recommended range or acceptable range, they had 12, 17 and 25 bushel yield penalties, or they would have, if you would have applied 144 pounds or the maximum of the acceptable range.
Jeff Vetsch:
And like Matt said, when we look at corn on corn, we had 10 sites of those, those were so much more variable. We had four sites that the economic optimum was less than our MRTN acceptable range, two sites that fell within the range, and we had four sites that were greater. And those four sites that were greater, we're 41 to 90 pounds greater than 180, which is the maximum of the acceptable range. So why are we seeing so much variability in corn on corn?
Jeff Vetsch:
I think it gets back to what Dan touched on and a little bit of what Fabian touched on. It's clearly precipitation is having some to do with it. I think these wet falls are also having something to do with it at. In these corn on corn sites. It was really interesting that of the 10, seven of them had very wet falls. Late September, October and into November on occasion and not every month, but some months where we saw anywhere from 150 to greater than, or up to 300% above normal rainfall or greater than normal rainfall. So that's really a major factor that could be driving how this corn on corn is behaving compared to corn after soybean.
Paul McDivitt:
So going forward, where does the group kind of see the nitrogen recommendations going from here?
Dan Kaiser:
Well certainly, when we started looking at a lot of our data, and again, we know that things are trending upwards for the rate response trials. The question, when we were looking at the numbers, what's kind of a more of the struggle was figuring out how much old data do we deal with versus how much new data. And that's been the challenge particularly if we see things changing. And the other thing too, is looking at more regional specific recommendations, because if you start looking at the numbers, can we come up with something that's more consistent within a particular region, one versus other that may be consistently higher one that may be consistently lower because that's mainly the challenge that we see, because we know there's certainly circumstances out there where the N rate response is higher, particularly for corn on corn.
Dan Kaiser:
But then I've seen other circumstances where we gets 250, 260-70 bushel yields with around 160, 165 pounds of N for corn on corn. So that's the issue really, is that variability that's out there. And then how do we address that? So if we start looking at the trials like Matt has in place, while we can't necessarily come up with the optimal rate with those particular fields, we can start looking at potentially with the number, if you have enough trials out there, being able to partition out certain areas that maybe we need to throw some additional N rate response trials, and then look at, or look at just the percentage of time that we know that we might be above or below. Because if we start looking at our MRTN approach, the way the database works, it doesn't generate an average of all the trials that are in there.
Dan Kaiser:
And generally, if we look at the actual response that we get for the recommendation, it's typically represented from around 75 to 85% of the trials where we've been able to be at, or slightly above the optimal rate for those particular trials. So, it's nice having some of this at least to know, because I think that's the main thing is really where do we target and where do we move forward, particularly with some of that data.
Dan Kaiser:
And looking at it on a widespread scale, a lot of these on-farm trials, it's nice for us because the traditional small plot trial, particularly for nitrogen rate trial, isn't the easiest thing when it comes to funding to get into place, just because it's not the sexiest thing out there in terms of research. So, there's certainly things we can do, particularly coupling these on farm to really look at this more and get a handle on where we need to be tweaking or fine tuning some of our recommendations.
Matt Wiebers:
To add to that Dan, I think the farmers and their advisors have really enjoyed this opportunity because if we look at on-farm research, it's not the easiest thing to do. It might look easy on paper, but in the heat of the season, a lot of times these things get skipped or overlooked or not done correctly. So with an advisor or somebody directing the grower, I think we can do some high quality trials. That's one of the things we've learned from this AFREC study is that I think the farmers are equipped to do this.
Matt Wiebers:
Not everybody's going to give us perfect data, but we have mechanisms to sort through the good data, and some of these guys are ready for a challenge. So I think if there's one thing that we could take away from this, I think it's that if we came up with a higher bar for some of these growers to do more intense on-farm research studies with more rate replications, I think they'd be willing to do it, and it would help everybody's knowledge base, maybe grow a little faster because these growers are willing to adopt things that they see on their farms. Maybe even after just one or two years, as opposed to looking at data basis from other parts of the state or other states where they might be a little more skeptical of those things.
Fabian Fernandez:
I think that, that's very important when we talk about these database, it really works on the number of sites. And as Dan was mentioning the part where we have to figure out how many of the old sites we give versus new sites. And the challenge is, if you don't have a lot of new sites, it's really difficult to take some of the old sites, because for instance, and as we've seen it, one year where we had a huge amount of rain, excessively wet, where we lose a lot of nitrogen and the response curves that are really high, that can really skew the data set a lot. And so the more years and sites that you have into the database, the better it is because he buffers those really radical years in the data set.
Fabian Fernandez:
And so I think what you said Matt is very important that we need to find more of these places, farmers, and advisers that are interested in generating some of these data. That are interested in going above and beyond the plus, minus studies and really finding more rates and really getting full response curves that then that information can feed into. It can feed obviously into their farming system. They will have very nice information for their particular farm, but it also contributes to the larger database. And then once you have that, if you have enough of that data, then we can start doing more of what Dan was mentioning in terms of partitioning the data and seeing if a particular region of the state needs different recommendations, or if particular soils or conditions require a different kind of approach in terms of how much nitrogen is needed.
Fabian Fernandez:
But until we have that information, it's very difficult. I mean, we've tried with what we have to see if we can partition the database and there is not enough information there to really do that. So, I think that, that's definitely the way that we need to go. And as we were also talking about the fact that in terms of resources, yes, the more people that we can have contributing to the database the better it is. And obviously, there will be some bad data here and there, but we can always find those and remove those out of the database or not include them in the database. But the key I think is to get more and more of those site years of data in the database.
Jeff Vetsch:
I would add that getting that local involvement and participation is important. We feel that the study that Kevin Keener was one of my partners at department of Ag. And when he came to me with this idea of working on these projects in Southeast Minnesota, our on-farm small plot and rate trials. He envisioned us working one-on-one with growers for a two or three year period, and trying to have more small group interaction and discussion with them. And then also getting their agriculture agronomy advisor involved, their retailer, whoever that might be. And we've done that.
Jeff Vetsch:
And I think that has been impactful. Not only do they see the results that we get and we share that with them, but when we meet with them in a smaller group, they feel like they have some ownership and they also have that identification and that we're taking that time to talk to them about what their needs are, and where they feel they could use this information moving forward. And ultimately, our goal is to try to get growers to have more confidence in the rate recommendations from the N rate calculator, and also adopt other BMPs for nitrogen that will help improve not only their economics, but also their environment.
Matt Wiebers:
You know, Paul, we talk about nitrogen on this podcast, but it is worth mentioning that the practices that the growers learned from doing nitrogen on-farm trials extends well into other nutrients, you know, phosphorus, sulfur, potassium, and then into other things that they might be interested on their farm, such as seeding rates or other crop protection things. So it does extend this. If we can teach growers how to do good on-farm research trials using their own equipment, I think we've empowered them to do things well beyond the scope of just what we're talking about today with nitrogen rates.
Paul McDivitt:
The next question is kind of about the future of on-farm research, Matt, could you talk a little bit about how precision Ag was incorporated in your project?
Matt Wiebers:
Right. 20 years ago, around 2002 or so, I started doing on-farm research down in Illinois. When I lived down there and worked with some of the faculty members who were at the university of Minnesota and as well as University of Illinois at that time. The technology has gotten better, but the practices really haven't changed that much. It's to start with growers that have the ability to soil test, to have the ability to record where those fertilizer rates are being applied, because we might plan to do a study, but what actually happens in the heat of the battle or once we get into the season, sometimes deviates from that plan a little bit. So we need to record what actually happens.
Matt Wiebers:
So collecting the as applied maps is a critical part of this AFREC project. And then of course taking some aerial photos in season to make sure that there's nothing out there that has impacted our study such as weather, too much rain, not enough disease, weed issues, things like that. And that's easy to do, there are so many satellites flying around now that we don't need to get in our four wheelers and drive out to look at these trials as much as we used to. I think that satellites and UAVs can take pictures of these fields and give us a good idea of what's out there at a low cost.
Matt Wiebers:
And then of course, carrying these trials through to yield, we know that a lot of these growers have calibrated GPS yield monitors that can give us high quality yield data. You know, if I look at the CV, which is a measure of statistical variability within these trials, in 2018, we were around 3% across all of our studies. The number went up to about 5% in 2019. And I think that's driven by the excess rain that we had impacting our plots and perhaps also the higher nitrogen rates as well. But those are some of the key things that I think about when approaching growers for doing on-farm studies and how we're going to get better in the future in doing these.
Dan Kaiser:
Yeah. And Matt, I know you had some comments just about the fact that, we shouldn't approach this to saying that this is easy, because to get quality data, you really need to put some time into it. And I know one of the biggest issues that we always run into too, is that a lot of times we have the growers willing and we have a plan set up, but the problem is the grower's usually working with a co-op, which then usually has an employee that's doing it. So you're a little disconnected in terms of who you've talked to that about doing it, versus who actually goes out and makes the application. So that's why those as applied maps are critical. I mean, that way we can go back in and even if there is some sort of a screw up in there, if it isn't major, we can at least know what happened.
Dan Kaiser:
And that's one of the things that I've found just with some of the stuff that we're doing, I've been working for a few years with the department of Ag with these nutrient management initiative, what they call a level two plot, which are essentially strip trials which we vary the nitrogen rate to get six nitrogen rates replicated three times within a field, just to try to get some sort of N response curve. And we've learned quite a bit from these, as applied to you again is critical.
Dan Kaiser:
The other thing too is, with just with the speed that things are getting applied out there too. There's just some issues out there, and so there's some things that have to be really watched out when you're looking at these, because again, you can do a lot of things, but it just taking the time to do it correctly is very important to make sure you're getting a good data set. And that's kind of the main thing with it. And certainly nitrogen is one of the easier things to work on and potassium as well, just because many of the fertilizer sources contain only those nutrients.
Dan Kaiser:
And when you start getting into more complex designs, phosphorous, even sulfur, if you have products that contain both nitrogen and sulfur, you need to make sure that you sit down and think about your project, that you have an apples to apples comparison, because we know that particularly out there, nitrogen and sulfur too that can interact, and also they both can be deficient. So if you're changing those variables and they're being changed independently, where you're not evening out, one of them, you can have some issues with it.
Dan Kaiser:
So it's said, take your time, have a plan, the good time of the year for us, that we like doing a lot of things here is over the winter, when growers have time, because I don't like having to do things in the fly, particularly in the spring. It gets done at certain points there, but it's just nice to have some time to think things out and have a good plan as we move forward.
Fabian Fernandez:
I think that, that's key too, to have good communication between all the parties involved, right? Because if there is a disconnect between the plan and who actually applies the nutrients or whatever is happening in the field, there is always chance for problems. And even in doing a small plot research, it's very easy to make mistakes when you try to make a decision that you need to make a change, you are in the field, you need to make a change, you have the plan, but something went wrong or something had to make a change. It's pretty easy to end up making a mistake because you're trying to do it on the fly.
Fabian Fernandez:
And maybe you don't remember all the details or why you were doing certain things the way that you were doing it. And so I think that's key to have a good plan of communication, and to have the commitment from the various groups to work together on these things that while it's being done in commercial farms, to a large degree is research and it needs to be treated a little bit differently than you would bulk areas of a feel when your farming as a bulk area.
Fabian Fernandez:
The other thing that I wanted to mention too is, sometimes there's this discussion about, "Well, what's better on farm research on kind of strip trials or a small plot research?" I get that question asked quite a bit, and I think both are very, very important. Both have a place and both are needed. We cannot just go with one and forget the other because they contribute completely different things. And they actually are synergistic.
Fabian Fernandez:
Some of the things that you learn in small plots you will not be able to learn in large plots because of variability and because of other issues, because you're limited in the number of treatments and variables that you can be looking at one time to compare things, apples to apples and things like that. So, it is important to do both. The small plot and the strip trials, but always having that communication is key. It's a key piece.
Jeff Vetsch:
Yeah, I just kind of add to what Dan and Fabian brought up about the biggest challenge with on-farm strip trials is you have to have a fairly limited number of treatments. So growers always are interested about when it comes to soil fertility, well, I want more than just one nutrient, I want to look at more things. I don't just put on nitrogen, I put on NPKS and something else. But trying to integrate that into one study on a farm and strip trials, this isn't going to work. So you have to kind of be simple, but make an objective. Know what you exactly question you want to answer and just do something small and simple enough that'll address that question without making it too complicated.
Dan Kaiser:
And when I look at it too, you look at these things you know, Matt, like the stuff that you're doing, the nice thing about that is, you've got a large number of trials. And that's one of the things that, you know you have a number of growers that have this similar type of question. There's some power in that, particularly if you can get a bunch of people together and get something out in the field, because then you can start looking at probabilities.
Dan Kaiser:
And that's kind of the main thing when I start looking at just simple things like product testing. If we have 50 trials out there, everybody has the same treatments out there, then you can start looking at how things work and an example is like these biostimulant products, which a lot of people are interested in. But it takes a lot of information, and while small plot trials on our end, one of the nice things is we can look at more complex issues, we're not really looking at that at a large number of locations. We just physically can't right now, just with the amount of work that it involves.
Dan Kaiser:
So, there's things that can be done, but the coordination is key. And that's where having AFREC support, particularly a supportive position that can work as kind of a central coordinator. Like what Matt's been doing is really critical with that as at least having somebody that can keep up with the growers and gather that data and get it together and get the data out there because it's really turnaround is, I know as a lot of these growers, they want their information quickly because they're making decisions fairly quickly in terms of what they need to do for the next growing season.
Dan Kaiser:
So, a lot of challenges kind of with this, but there is some power in some of this data and particularly having a number of locations with the same treatments is really I think what's critical to look at getting enough data where you can make a more accurate decisions in terms of what's worth and what's not in terms of specific practices.
Fabian Fernandez:
I was just going to add a little bit to that. At first look, it almost looks like a brute approach, right? It's like more numbers, but as Dan mentioned, it's that probability trying to understand what is possible that the only way to really do that is to be able to look at probabilities. And so you have to have numbers. The more numbers you can have, the more side years of data that you can have, the better it is, because there is a lot of variability out there at the large scale, and even at the small scale. I was just looking at the data from an AFREC funded project where we are looking at in very small scale, basically about two acre size plots where we have within that two acre, we have six nitrogen response curves, and we can look at the MRTN values for each one of those response curves.
Fabian Fernandez:
And you see the difference within a two acre field, it's pretty substantial. The MRTN varies 40, 50 pounds, and you're only looking at two acres. So that variability is real, is there. And so, when you take these two home farm sites, that variability is there, and maybe more if it's a larger scale or larger in size. And so, having a large number of data points to really understand what's going on and be able to answer what is the probability, how many times out of 10 or whatever you get this nitrogen rate working is a critical piece of information. And the only way to obtain it is by having a large database.
Matt Wiebers:
I think one thing that Fabian's mentioning at this point is that what we're doing with farmers and their advisors is not meant to try to refute or dispute the university recommendations. I think on the contrary, it's meant to generate data that reinforces that. And if we go back in time to 2013, 2014, there was actually an AFREC funded project to look at how we would arrange this. And I think that was something that Carl Rosen, Mike Schmitt and John Lamb had done at the time. It's a partnership between university making recommendations on the science, and really what we need to do in the country as farmers to create good on-farm research trials.
Matt Wiebers:
And then, it's finding people who are advisors or influencers of those farmers to get three, four, five of your growers to actually do a couple of these studies. And if we could get 10 or 20 or 30 of these advisors around the state to do this every year, think about the hundreds of trials that we could come up with in short order, it would just really allow us to scale these up without a lot of investment or effort.
Dan Kaiser:
And that's one of the things that I see. I mean, I know our recommendation system isn't perfect. I know we don't hit the exact MRTN then for a particular field all the time. Matt, I agree with you on that. Having the capacity to scale these things up has a lot of power because a lot of people, particularly will get into the regulatory realm while our recommendations aren't meant to be regulation. It does work into that way just because of some of the government payments in the programs in terms of what these growers have to use.
Dan Kaiser:
The more information that we have, the better than that we can go in and then hopefully look at fine tuning these things and coming up with some sort of systems, particularly if there are situations where you're seeing consistently a higher nitrogen requirement, is there something that we need to be looking at specifically in that area to make a specific recommendation, it's something that we can do. And that's one of the things that, again, I struggle with because as Fabian mentioned earlier, we did look at partitioning our recommendations out to try to be more specific in some of the areas, but it's really a numbers game.
Dan Kaiser:
And that's one thing I know Fabian coming from Illinois, you said Emerson NAFZiger, that’s one of the things they did when the MRTN in place was to get a lot of trials out there, to start looking at consistency across the state. And they have multiple recommendations based on where you're at in the state. Right now in Minnesota, the areas that I'm looking at separate, sands we know higher nitrogen requirement. And then we know there's some differences up in the Red River Valley. But beyond that, I said, it's been hit or miss.
Dan Kaiser:
So, as we go along, particularly with corn on corn is the area that we've less information on. And now with Fabian starting a few years back, we've been able to get more data. We've been able to increase some of our database but we could always use more. And that's kind of the main thing. So that's really the avenues we're looking at. It's really the best way to go about this.
Paul McDivitt:
All right. Any last words from the group?
Matt Wiebers:
I'll say this, I think that there's a lot of people who will listen to this podcast who probably haven't done an on-farm research trial or who may be interested in doing one, but not sure how to get started. I think one of the key things for advice that I could give you as a listener is collect good data on your farm, even if you're not going to use it for an on-farm trial this year. A lot of the on-farm trials that we designed, we want to go back in time. We want to look at the yields last year, two years ago, five years ago, and use that data to help design a trial.
Matt Wiebers:
So, calibrate your yield monitor and make sure that you're collecting good data, because when it does come time to do an on-farm research trial with precision ag, we're going to ask for that data. And if that data is high quality, it's just going to make your trial all that much better.
Dan Kaiser:
The other thing I'd bring up too, and I didn't mention it before was replication too. That's important and that's one of the things that I know if you talked to a lot of consultants and people that we work with, they understand this, but there's times with farmers where I know it's difficult when you start looking at trying to replicate some of these things in the field. But if we're trying to tease out some of that within field variability, because you look at a yield map, your yield map, isn't the same color across that entire field.
Dan Kaiser:
So the only way that we can try to tease out some of that variability is with replication. And it's really important that you start looking at that. And one of the questions that always comes up is can I just increase my strip length or can I go to a wider strip to not have to replicate, but it just doesn't work that way. So, there are ways I know statistically that people can do looking at, if you have a side-by-side comparison of along a number of fields where you can feel this replication, but really the gold standard is to try to be able to replicate things in your field.
Dan Kaiser:
So, I think the thing to stress on that is, again, planning is really important. Get these things set up, get them set up right, to make sure that you can get the data out of it that helps you make a decision. That's ultimately what you're trying to do so you're not wasting your time and spending a lot of time, effort and money on something that you're really going to be looking at bad data.
Fabian Fernandez:
And it can feel a little daunting if you haven't done these things. When we start thinking about replications and treatments and all these things, but in reality, it's not that complicated as we were talking earlier. The technology is there for us to do these things actually fairly easily. It does take a little bit of planning, but that's where a lot of us have those expertise in terms of developing a treatment plan that you can apply and then using the technology, you can do that fairly easily.
Fabian Fernandez:
And then as we were kind of talking about it, the important thing is once you've decided that you want to do this to make sure that you put your eyes on that as part that you take care of it. That you pay attention to what's happening because it's easy to make some mistakes if you treat it just as the rest of the farm kind of thing as a whole field type approach. But it's not really that complicated to do it, and there are a number of resources at the university and other places where we can help set up some of those trials if you're interested.
Dan Kaiser:
Yeah, and I will say that auto guidance is a wonderful thing when it comes to some of these things, particularly if you can skip passes, but I know it said it takes time and in the heat of things, we know our season, isn't all that long and to get things into the field is really the important thing. So that's kind of the main thing with it. You've got to have the time to do it, and to do it correctly. So our technology is there, we know that technology is solid. It's just how we use that technology is really what we have to be careful with.
Jeff Vetsch:
The last thing that I would mention is it kind of goes back to the recommendations that people give for choosing soybean varieties or corn hybrids. Don't overreact to one site one year wonder. You need to look at things over time, and you need to look at it more than one location, more than one farm, more than one site year, more than one environment. And don't overreact to just one number or one experience that you might have.
Paul McDivitt:
All right, well that about does it for the podcast this week. We'd like to thank the Agricultural Fertilizer Research and Education Council, AFREC for supporting this podcast. Thanks for listening.
(Music)