0:05
OK, Hey everyone.

0:08
So what I’m going to talk about today when I so I was thinking, OK, let’s talk a little bit about nitrogen.

0:12
Let’s talk about, let’s talk about working with the data at the back end, right.

0:17
So everyone’s interested in maybe running on farm nitrogen rate trials.

0:20
We have a lot of programs that are happening right now in the state to support that.

0:25
We have a big initiative, I mean, not we have this, you know, a big on farm research program as well.

0:31
But let’s talk about some of the finer points of you’ve run a nitrogen rate trial and now what do you do?

0:38
So I’ve got a presentation here and co-authored with Monica Schauer, who’s doing a lot of work with the with the nitrogen optimization pilot program from DATCP.

0:53
So a couple things.

0:56
So I’m going to go through some components of a nitrogen response curve.

1:01
I want to talk about some and show some examples of ways to evaluate that economic optimum N rate and I want to kind of touch on that a little bit more.

1:10
We talk about that economic optimum N rate, but we got to talk about what it really means.

1:13
I want to talk about then a few of sort of maybe the pitfalls or the problems you might encounter specifically with flat, relatively flat curves and problems when rates don’t go low enough.

1:25
So the examples that I’ll show here today are ones that from research trials that have been conducted over the past few growing seasons.

1:32
So we’re going to start also with this kind of premise, right?

1:34
So this is going to be about running an on farm nitrogen rate trial because they can be a valuable tool for farmers to identify optimum N rates as a way for us crop consultants, agronomists to engage scientifically with our clients.

1:49
All right, it’s a great way.

1:51
It’s relatively simple.

1:52
It’s a great way to kind of get in on, on farm research.

1:57
And, and it’s, and it’s quite popular, right?

2:01
I think that these nitrogen response curves can provide a lot of insight and for that field in that year, right.

2:07
And, and for that farmer.

2:09
And what I want to do is kind of go through the, the approach of how, you know, I, how I work with nitrogen response curves and what I think about when I look at the response curve itself.

2:18
I also want to be very clear, this is not about building a nitrogen recommendation system.

2:22
This is about conducting an on farm research trial and gaining as much information from that to give back to the farmer.

2:31
OK, so, and then as I talk, it’d be great.

2:33
Yeah.

2:33
I don’t see the chat, but if there’s any questions in the chat or if anyone wants to stop me, it’s always better to, you know, answer questions throughout.

2:42
So if anyone has any questions throughout, if I’m, if I skipped over something, I’m not explaining something, well, let me know.

2:48
All right, so here we go.

2:50
All right, this is, this is a nitrogen response curve, right?

2:53
And what are the components of it, right.

2:55
When we talk about this?

2:56
So I’ve got, you know, I’ve got these little numbers here to remind me all the things I need to point out here.

3:01
So number one, we have the X axis.

3:03
It’s the different rates of nitrogen applied.

3:05
So here is a case where the rate started at the minimum rate was 50 and went up to 200 lbs in 40-LB increments or some along those lines.

3:20
In any case.

3:20
So, and then if we have multiple reps in there, right?

3:23
So we have multiple plots at each, at each nitrogen rate 2, you got the X axis.

3:28
This is yield.

3:28
This is corn yield and bushels per acre.

3:30
What you’d measure on each plot #3 is the curve.

3:35
And this is really what we, what we’re after, right?

3:37
We want to build a response curve.

3:39
And this is a case where the response curve is what we call a quadratic plateau, meaning this gets in sort of the aspect of law of diminishing returns. Nitrogen gets applied and yields go up to a point.

3:54
And then at that point, yields don’t start to go back down again.

3:56
We’re not working in any sort of toxic levels of of nitrogen.

4:01
It’s just yields don’t increase, right.

4:03
So the sort of response function that generally best fits most response curves is this idea of a quadratic plateau.

4:11
It, it increases with a quadratic function and to a certain point, then it flattens out.

4:17
And that’s all that means #4 over here is the R-squared of the response curve.

4:23
So how how good of a fit is the curve relative to the data?

4:28
Another way to think about it.

4:29
So this has an R-squared of over .9.

4:31
This is the, it means it’s a really, it’s a really good model and implies that 93% of the variation in yield can be explained by the nitrogen rates themselves.

4:44
So there’s not a lot of variation leftover.

4:46
So it’s a, it’s a really strong response curve.

4:49
Then from those response curves, we can produce 2 pieces of information and we’re going to use a lot of acronyms here, but the AONR, that’s the agronomic optimum nitrogen rate.

4:59
That’s the rate at which yields plateau.

5:02
So in this case it was 100 and at 160 lbs of nitrogen.

5:06
However, there’s again with this idea of the law of diminishing returns and this economic optimum rate, this EONR, there’s a level at which it wasn’t worth applying.

5:18
OK, here, let’s start with this.

5:20
The EONR was 154.

5:22
That’s the rate at which the return to nitrogen was maximized.

5:26
It’s the rate at which the most profitability occurred, and that was 6 lbs less than the agronomic optimum nitrogen rate.

5:36
So again, this is a really tight curve.

5:38
The number of the vows are very close to each other.

5:41
But so it implies like, OK, it wasn’t worth applying that extra 6 lbs of nitrogen, you didn’t make any make any money on it.

5:48
All right, so far so good.

5:50
OK.

5:52
But when we talk about this economic optimum nitrogen rate, so if we say it for this farmer, the most profitable rate would have been 154 lbs of nitrogen.

6:01
But there’s a, there’s a range of values there, right?

6:04
It’s not like if you applied 1 LB more or 2 lbs less like you’ve lost a ton of money.

6:12
So we often use this, this idea of, well, what’s the range in that optimum rate to manage within about a dollar per acre net return.

6:23
And that rate goes, it’s so we have again that optimum rate of 154.

6:27
But the really that the range of the rates that are within a buck.

6:31
And we kind of use that as a, you know, OK, we’re, we can manage and you know, within a buck is, you know, no one’s, no one’s probably really going to care about a dollar on either side of that, right.

6:40
So that is more of a range of 150 to almost 160.

6:44
So, it’s about a 10 LB range you could apply in where you’re, you’re pretty close to maximizing your profitability.

6:53
What’s a little interesting that I want to point out and this, the, this response curve is very tight and it drops off pretty rapidly.

6:59
So, this range isn’t very necessary very large.

7:03
But also it’s also important to note that this range is not equally distributed going down the curve versus going up the curve.

7:11
So, if you’re going down the curve, right, so if you start at 154 and you start to go down the curve, you can only go 4 lbs less before you aren’t maximizing your return within a dollar, right?

7:27
And that’s because then the yield starts going down.

7:29
So the price of corn is relatively valuable to the price of nitrogen.

7:33
So you can’t really go that far to the left.

7:37
You can go a little bit further to the right.

7:38
So here we go up to it’s 5 lbs of nitrogen.

7:41
It’s still kind of tight, but it’s, these trends work out throughout all the response curves we’re going to look at.

7:49
You have a wider range of going to the right.

7:51
Why?

7:52
Because nitrogen isn’t that expensive relative to the price of corn.

7:55
You didn’t add any.

7:56
You’re not making any more money on your corn, but the price of nitrogen isn’t, didn’t really cost you too much.

8:05
You’re not losing too much money by applying an extra 5 lbs of nitrogen.

8:08
So anyway, we’ve got a range of 150 to say 160 where that’s kind of the profitable range.

8:15
All right, everyone good so far.

8:17
All right, we’re having fun looking at data.

8:19
All right, more data, what things can change though, right.

8:23
And so we can use these sort of curves with, and we can also apply different economic scenarios.

8:28
Like historically we use this price ratio of .1.

8:32
So assuming about 40, you know, the pounds of 40 cent nitrogen for a pound of nitrogen $0.40, roughly $4.00 corn, right?

8:40
That would be the price ratio of .1.

8:42
However, you know, right now nitrogen might be a little more expensive to to corn.

8:46
Maybe the price ratio goes up a little bit.

8:48
Maybe nitrogen gets a little more expensive to corn.

8:50
Maybe you want to have you want to run some scenarios.

8:53
Well, now in this like really good response curve that we’re starting with, you know what you know, also don’t hope that you’ll get a curve like this.

9:02
This is don’t get your hopes up, but this curves like this could happen.

9:05
Really good curves can happen.

9:07
It was only about a dollar or I’m sorry, it’s only about 1 LB of nitrogen difference.

9:11
Even if even if nitrogen got a little more expensive, the economic N rate didn’t really change that much in this in this scenario.

9:18
And that’s nice to notice and it’s nice to know.

9:21
And the reason why we know it is because of the curve we generated, right?

9:24
So having these nitrogen, different nitrogen rates spaced out and then we can use regression to draw the line.

9:32
And so I think then this is a case where it’s very helpful to have a really nice curve and you can apply different economic scenarios, all right.

9:42
But we can also take a look at, well, you know, OK, that’s within a buck.

9:46
What about $2 or $5 per acre range.

9:49
What is, what are sort of these numbers look like?

9:52
And you can see like, obviously when you’re managing, OK, well, what, what if, what if you want to maintain your profitability within $2.00 per acre, the range goes up a little bit, Not that much though, right?

10:02
So it went from 1:50 to 1:59 to 1:48 to 161.

10:06
So just a few extra pounds on either end.

10:10
What if you were a little, you didn’t mind being profitable within $5 per acre in terms of relative to the maximum of economic return, right?

10:21
If you can manage within $5 of economic maximum, well that range gets a lot bigger.

10:27
So you could apply as little as 144 or as much as 167.

10:31
And then we also put, you know, OK, $10.

10:34
I think that’s probably a threshold people would probably certainly want to be below, you know, within $10.

10:39
So you really probably $5 might be a maximum return that farmers may want to think about.

10:45
But you can see that range, the range gets bigger and bigger and that range got a lot bigger on the right side, right?

10:53
So it went down to 144, but it, it’s going up faster on, on, on this side.

10:58
So you anyway.

10:59
So it could be 145 though, almost 170 lbs of nitrogen.

11:04
So you can see they’re like, there’s really a range because of the price of nitrogen.

11:08
Optimizing economic return really is a range of values, not just a single value.

11:15
And you can play around with all of this to think about how close do you want to get or how, how dialed in do you need to be.

11:22
OK, all right, so in a nitrogen response curve like this, what did we learn?

11:28
We learned that the economic optimum rate was 154 lbs of nitrogen.

11:32
We learned that this was a very sharp response curve.

11:35
So the there was a relatively efficient use of nitrogen.

11:38
So, as you’re increasing your nitrogen rates, it’s likely that that nitrogen is getting utilized.

11:44
So it’s nice to see that sharp response curve, you know, a low range of economic return, even up to $5 per acre that that range was still relatively tight.

11:54
So that’s again with this nice curve.

11:58
And we learned that, hey, this farmer typically applies 160 lbs of nitrogen.

12:02
Hey, that’s pretty good.

12:03
I think they nailed it, right?

12:04
Excellent.

12:05
So if farmers see this and be like, hey, everything, I’m doing a good job.

12:09
No need to maybe change anything.

12:12
But there might be other pieces of information that we need to know, right?

12:15
Were there any manure or alfalfa credits in this field?

12:18
There were not.

12:19
OK, so we don’t have to worry about anything along those lines.

12:22
Did it follow a cover crop?

12:24
Now in this case in this field it did, but when we looked at how much cover crop biomass there was it was way below a threshold at which we would it was below 1000 lbs per acre, which is typically a threshold I like to use before that cover crop starts messing with the optimum nitrogen rates.

12:43
So there was cover crop used but it wasn’t big enough.

12:46
So OK, no need to factor that in.

12:48
Were there any weird weather patterns?

12:50
Yes, it was a wet spring, but the farmer applied the nitrogen at a side dress timing at this at this rate.

12:59
So that probably negated any bizarre weather patterns because it it was applied a little bit later into the into the growing season.

13:07
And then of course, if any, if anyone’s collecting pre plant or pre-sidedress nitrate test in this case we didn’t have any, any samples for this field.

13:16
So we don’t have to consider that in this case.

13:18
So those are a lot of the other things that that you want to consider.

13:22
You want to have all of these sort of, you want to ask the former, you know, all of these pieces of information and we have some standardized protocols if you’re interested on, on maybe what should be asked.

13:33
I’m hoping if I’m hoping my time’s out, somebody’s just going to yell at me.

13:38
All right, So the other thing is like, well, why use regression instead of ANOVA, right?

13:46
Everyone learns ANOVA where you’re just we’re interested in comparing the yields at each nitrogen rate, right?

13:53
And we can say, Hey, was there a difference?

13:56
Now, the difference between these two, the whole point of putting these nitrogen, using multiple nitrogen rates, 5 or more in most cases is what I like to see.

14:06
I love six. Really love 8. 5,

14:08
I can handle sometimes.

14:11
But the whole point is with the, when you’re using these, the whole point of doing it is to generate the curve.

14:18
And then the value of having that curve means you can make inferences between all of the exact end rates that you put in the study, right?

14:25
We have this line.

14:27
So you can make inferences of where yield would be and then what economic return could be at any nitrogen rate.

14:33
When you’re running analysis of variance, you’re only interested in the specific nitrogen rate used.

14:39
Now, this is a case where both the ANOVA and the regression actually work together ’cause you could see that here we’re saying, well, yields were maximized at 100 and you know, 160 or, sorry, at 160 lbs we got 186, at 200 lbs we got 186.

14:56
Those were the same below that.

14:59
Once you went go to 120, yields were statistically reduced, right?

15:04
So, OK, so if we use ANOVA, we would have come to the same result.

15:08
But for most cases that that may not be true in most cases.

15:12
So you can certainly use an analysis of variance and take a look at that as like a backup.

15:17
But really what you want to do is, is develop the response function.

15:23
All right, so now let’s go into something that’s a little trickier, right?

15:26
We have this, you know, this flat curve problem, or at least potential problem.

15:31
So this is, this is another field by the same farmer, but a different field.

15:37
And we were, there were six nitrogen rates.

15:41
I got this one was nice and started at actually.

15:43
It actually started at 0.

15:46
But then you can see that response curve is a lot flatter and you can see that R-squared is a lot weaker, right?

15:53
It’s, it was an R-squared of .36 only only 36% of the variation yield is driven by the the nitrogen rate.

16:04
But in any case, we can still fit a response function here.

16:07
A quadratic plateau still did fit and we found an economic optimum return.

16:11
So we maximize yield at 85.

16:13
No, excuse me, we maximize profitability at 85 lbs of nitrogen.

16:17
OK, so that’s a lot less than that other field, right?

16:21
So what’s going on here?

16:22
Well, a couple of things before we go into what’s going on here, let’s take a look at when we’re playing around with the, the economic return.

16:28
So when we, you know the, the first thing I didn’t put the numbers on here, but this is that the one on the left is that classic .1 ratio where it’d be 85.

16:37
and then on the right we see an economic return or economic optimum nitrate of 76 when the price of nitrogen goes up a little bit at that .125.

16:49
So now in the other one, remember there’s only a 1 LB difference as the price of nitrogen up went up based on this curve.

16:57
If the price of nitrogen goes up, you want to reduce by 10 lbs.

17:01
So it has a much bigger of the flatter curves.

17:03
There’ll be much bigger effects of that price ratio on flatter curves and there’s going to be a bigger, you know, the bigger and bigger range.

17:14
So like if we’re we’re doing the same thing with all these differences and economic return, let’s look at this $5 range.

17:22
Well, start with the, the $1.00 range.

17:23
So at 85 it would go 79 to 97.

17:27
So you know, 80 to almost 100.

17:29
That’s almost, you know, it’s almost a 20 lbs range.

17:32
And then as you go to maybe like a $5, if you want to maximize your profitability within $5 an acre and there you’re going 60 to over 110.

17:43
So it would be in a bigger range.

17:44
So these flatter curves, these ranges get bigger and bigger.

17:48
So that’s the other aspect of of these flat curves.

17:51
So sharp curves, everything is nice and tight.

17:53
These flat curves, it’s a much bigger range.

17:56
And it does demonstrate some of the some of the issues you know, of like the economics of, you know, you can apply.

18:03
You know, this farmer could have applied almost 100 lbs could have and still been, you know, quite profitable, even though there’s very little yield gain to be to be made, right.

18:15
So that’s why I mean, that demonstrates why a lot of farmers are more willing to err on this side of the curve, the right side of the curve than the left side of the curve.

18:23
So that’s a big aspect of this EONR.

18:27
It’s the distribution on the left and right aren’t, aren’t equal in terms of risk.

18:35
All right. So, what else is going on here?

18:38
Why 85? Fairly low, right? Response fairly flat.

18:42
Well here we go.

18:42
Previous crops this was a this site was in 2023. 2022 crop was corn. 2021 alfalfa.

18:50
OK. 50 LB nitrogen credit from alfalfa. So, that expected explain some of the reduction.

18:58
We also the site also collected because the nitrogen was applied at sidedress, a pre-sidedress nitrate test was collected.

19:05
Got a value of OOP.

19:06
That’s, I’m sorry, that’s not 18 lbs.

19:09
That should be 18 part per million and that would suggest a 100-lb nitrogen credit.

19:16
OK, so let’s work through this.

19:18
So if the farmer typically applies 160 lbs to this field and typically does not take the second year credit, what do we learn?

19:24
We learn that hey, you should take the credit.

19:27
There was some value in taking that credit, but what we didn’t quite dial in was which credit to take.

19:35
So, they start with 160 lbs and you subtract 50 from that for the second-year credit, we’d say the optimal end rate was 110.

19:43
So that would be 25 lbs greater than the optimum rate.

19:50
But if you use the pre-sidedress which suggests 100-LB nitrogen credit. 160 -100 is 60. The optimum nitrogen rate then would be suggested at 60.

20:04
Well, that’s 25 lbs less than that.

20:06
So, we’re 25, each of the credits put it at 25 lbs above and below what the actual optimum rate was.

20:12
So, we didn’t learn which crediting thing was better.

20:18
So in most cases, you know, I suppose more likely to err on the lower value for the credit.

20:27
And again, you know, farmers are more, you know, it’s it’s there’s there’s less risk of being on the high end of nitrogen rates compared to the low end across that range.

20:39
OK.

20:39
Is that all make sense?

20:40
So that one’s a very specific case where we had a lot of information on the from that from that system.

20:49
All right, I’m, I’m assuming I have a few more minutes.

20:54
No one’s yelling at me yet.

20:57
Yeah, you have a few more minutes.

20:58
Cool.

20:59
All right.

21:00
So, now here’s one more type of scenario where you maybe you want to run nitrogen response curves to test something you have you want to run nitrogen rates on different parts of a field to know if, if, if, if that if, if the nitrogen optimum nitrogen rate varied in by landscape position or drainage class.

21:21
In this case, this is the case where we put out a biological product and we wanted to know if this product was the optimum nitrogen rate would change if you applied this biological product or not.

21:33
So there’s a couple of things I want to talk about.

21:34
So we’ve got 2 curves on here, right?

21:36
One is with the biological product and 1 a nitrogen response curve without the biological product.

21:42
But there’s a couple of issues that are happening here.

21:44
So one or one things we want to talk about 1 is talk about comparing the curves and the other aspect is a case where because on this site 100 lbs of nitrogen went out with late season fertigation.

21:58
The minimum rate in the study really is 100, right.

22:02
And so the when they varied the sidedress application, which they did across 5 rates, still the minimum amount of nitrogen went out was 100 lbs.

22:10
And is that low enough?

22:12
OK, so let’s talk about like comparing the curves.

22:16
So when we look at something like this, we’re looking at an individual curve.

22:19
And you may say that, well, this looks different, right?

22:22
The curves, I see a visual separation of the curves, but there’s a lot of variation, right?

22:28
You can look at the data itself.

22:29
There’s a little bit of spread there.

22:31
So what you really want to think about when you’re, you want to think about comparing the curves, but you want to think about comparing them kind of with an error bar.

22:39
And so you can kind of see for the red curve with this reddish sort of shading that is the, that’s the 95% confidence interval.

22:48
The Gray curve has the Gray, you can see those confidence intervals really are lying right on top of each other.

22:54
And so it’s important to graph stuff like this and look at stuff like this to know like these curves aren’t different.

23:01
Nothing is different about these products and how they’re influencing the optimal nitrogen rate.

23:10
I’m going to just skip over that for now.

23:13
The other thing on this was the that we ran into the big issue was that the the curve is relatively flat and with a relatively flat curve and the idea that we don’t have rates less than 100, we don’t have a drop off at the back end.

23:31
And so without that we kind of struggle with where the optimum nitrogen rate is.

23:37
So based on this response function, we couldn’t based on the regression, the, you know, the, the potential optimum nitrogen rate was at the highest rate that we use, which was 220.

23:49
Now we start to put, we start to put these, you know, within these dollar ranges or $5 ranges.

23:56
So in any with a with to manage within a dollar, you could drop down to, you know, anywhere between 214 to 206.

24:04
So anywhere between 5 to 10 lbs less nitrogen to still maximize to manage within a dollar per acre profitability within a $5 profitable range, you could drop down as low as 186 to 196.

24:19
So you could drop down almost 30 lbs right and still be, you know, still be within a, you know, a fairly profitable range, close to a profitable range.

24:29
So, so the problem with again with the flat curves, these error the the this range, this profitable range ranges are very wide, they’re very wide.

24:40
So we able, so this is a great, this was a nice study because we were able to go compare the products, right.

24:47
We know that there wasn’t an effect of that product. But we still like, if we want to know anything about like our optimum nitrogen rates that end up being a bit limited.

24:57
You know, it’s a case where they did use 5 rates, but they’re all within that part, that flat part of the curve.

25:03
So, you want it’d be better to have more nitrogen rates or that those five rates distributed across like the responsive range and the flat range, in addition to some of these sites are just with on from research and any plot level research too, you might have a fair amount of variability, right?

25:20
And that kind of maybe masked some of potential effects that we could have seen or used to identify optimum nitrogen rates.

25:27
So it’s okay.

25:28
I mean, sometimes in the case like this, this was the best that a farmer could do to run a nitrogen rate trial.

25:35
But it’s not going to be a case where it’s going to be that it’s not that informative for maybe for optimizing nitrogen, but it’s a good, it was a good approach to test different products.

25:49
All right, so, what are the take home messages?

25:54
I think that these on farm N rate trials are fantastic.

25:56
They can provide tremendous insight into that economic return and it knowing what the curve looks like, right?

26:03
I think that’s maybe the most important thing, right?

26:07
In that first case, it was like that was a really sharp curve and it really demonstrated that farmer would like had nailed their, you know, they’re working at their optimum N rate and they can feel good about that.

26:16
The second one we looked at is like, oh, they need to start thinking about crediting.

26:19
There’s some tremendous, there’s real value in that, right?

26:23
But you wouldn’t know that without the curve.

26:27
There’s a, there’s a ton of nuance in this, but we’re here to help.

26:32
All we do is run nitrogen response trials all day long and, and look at the results.

26:36
So we’ve seen a lot of stuff, but it takes a while.

26:39
You know, we’re with these NOPP trials.

26:41
We’re looking every single one.

26:42
And Monica and I meet and, and Abby’s part of that group and Lindsey’s part of that group.

26:47
And we sit and we look at and it takes us an hour just to talk through all the details of every different, all the different things that are happening in that field that might influence the the end result.

26:58
And then the last take home message is that I think everyone on this webinar should plan to do at least one farm N rate trial in 2025.

27:06
Well, let’s start them and plan to do it in 2025.

27:09
Let’s run them in 2026.

27:11
And then to put that in context, hey, there is funding out there for this, right?

27:16
Of if you’re working with crop consultants, if you are a crop consultant, you’re working with farmers, consider applying for these grants.

27:24
It’s, there’s a lot of, there’s a lot of funding out there.

27:28
We need, you know, there’s a lot of, there’s a lot of things that need more testing.

27:35
That idea of different soils within a field, landscape position, drainage class, what if you changed your timing of the fertilizer, change the source of the fertilizer.

27:47
All of these different things are incredibly interesting that we really can’t have.

27:51
There can’t really be enough nitrogen rate trials on.

27:53
So it’s a case where the farmer would learn a ton and then any sort of information out of that too would be really, really valuable.

28:01
So if anyone’s interested, Jerry’s got the link or just Google DATCP NOPP and take it to the application process.

28:10
If you’re also interested, don’t hesitate to reach out myself, Monica, Abby, Lindsay, really anyone we’ve had, there’s quite a few of us that have a lot of experience with helping design these types of trials and we can certainly be helpful in in for people that are submitting a grant.

28:31
So with that, I don’t know, were there any questions?

28:34
Hopefully people found that somewhat interesting, insightful.