The Sixth Sense

Jackson: Welcome to the FarmBits podcast, a product of Nebraska Extension Digital Agriculture, I'm Jackson Stansell.
Sam: And I'm Samantha Teten, and we come to you each week to discuss the trends, the realities, and the value of digital agriculture.

Jackson: Through interviews and panels with experts, producers and innovators from all sectors of digital technology, we hope that you step away from each episode with new practical knowledge of digital agriculture technology. Hello FarmBits followers and thank you for joining another episode of the FarmBits podcast.

Sam: This episode continues on our nitrogen management series particularly within the topic area of responsive nitrogen management.
Jack: Yeah, our guest for this episode is Dr. Josh McGrath who is an associate extension professor in soil management at the University of Kentucky.
Sam: Josh is a soil fertility specialist with significant experience in nitrogen management in both Kentucky and mid-atlantic region. Jack: Some of his research and extension efforts have involved exploring proximal sensors for informing in-season nitrogen management and small grains and corn production.
Sam: So, in this episode we will get into what proximal sensors are, how they are used, and what the results have been from integrating these sensors in commercial farming operations.
Jack: This episode reveals the realities and the potential of proximal sensing solutions for nitrogen management, so sit back and enjoy this conversation with Dr. Josh McGrath.
Josh: So, Maryland is very diverse and part of the state's not even in the Chesapeake bay watershed. The western two counties out in the Appalachian mountains, they're mountainous very little row crop and outside the Chesapeake bay watershed and then you go east and then you're on the western shore of the Chesapeake bay around Annapolis, DC, Baltimore and you get into that western shore coastal plain, and so those are historically eroded soils in southern Maryland. Their coastal plain but with rolling topography, somewhat depleted soils because of historic and when we say historic like you know 200 years of tobacco production and then you go keep traveling east across the bay and you're on the Delmarva peninsula. So, that's all of Delaware where I grew up. The eastern counties of Maryland and two counties of Virginia and so you get further south on that coastal plain and it's just beach sand and it's flat. We had a field site we were measuring runoff at, a ditch that we had surveyed and there was like one inch drop in two miles. I mean it's flat, it's like a pool table right and you're like three feet above sea level. So, you get salt water, tidal flow back up into some of those ditches you know, and so it's an interesting place to farm. So, you've got first of all you got diverse and that diversity leads to very different perceptions of the farmers and this is an important part when we talk about doing precision ag and variable rate. You know where I grew up on the Delmarva everybody had been split applying nitrogen on corn you know some up and starter two by two liquid fertilizer or broadcast nitrogen up front. You know maybe 50-100 pounds up front and then you're coming in with another 100 pounds just drag hoses on a big hundred foot boom sprayer, high clearance sprayer. Drag hose you know just dribbling UAN down the center of the row at that v4 to v6 kind of up to the 10 stage. That's across the board that's how everyone does nitrogen they're putting it through equipment lots of irrigation. You go west of the bay, and when I started in 2006 one of the first extension talks I ever gave was in Baltimore county so just north of Baltimore city, and it was mind-blowing to me I didn't know there were folks that didn't split apply nitrogen and no one 100 of people just put down all pre-plant, which is what as you move to the midwest people you know. But, I'm just a young guy, I knew what I knew right. I knew farming on Delmarva, and so it was wild to learn about people you're not split applying nitrogen because in nitrogen management like forget precision ag the the big knob the huge lever like if the older folks you you all might not remember course knob and fine knob on tuning the radio. But, the older folks will remember the core snob the big knob is tiny, but if you're not splitting applying forget about precision you're wasting your time. There's no point in doing variable rate nitrogen if you're not split applying because you're missing, you're working around the edges. You're worrying about the little tiny details, and you're missing the mark by a 100 miles. So, forget about everything split apply and then worry about getting into precision.
Jackson: We're kind of starting to shift this discussion into the precision ag space, and we've talked about how you know people that are out on the Delmarva are already split applying nitrogen to kind of take care of some of that efficiency, but with some of these challenges to try to get to that NUE that we talked about you have to kind of take that next step whether that's non-application through the pivot or whether that's going to a precision ag sort of application where you vary the rate across the field, and so what are some of the tools the precision ag tools that farmers are employing in that particular part of the country and even in Kentucky right now to kind of meet some of these challenges head on with nitrogen management?
Josh: Yeah so, I mean I think most of the nitrogen management it's funny, and I'm gonna kind of start talking about both nitrogen and phosphorus as a means of kind of comparison, so when we look at some of the survey work somewhere between 40 to 60 percent of the phosphorus applied to row crops in North America's applied variable rate and probably most satisfied we're guessing is based on some sort of grid sample and they're just taking our recommendations that existed before or whoever's recommendations and they're basically saying okay we were applying this phosphorus rate to this range of soil tests and now we have this gridded out interpolated soil map from this grid soil sample. Now, let's set aside whether or not that's right that interpolation is right let's assume we're going to let's say that you did a grid sample, you took a sample every every 60 feet a tenth of an acre so 10th acre grid sample, so we know it's right yeah and then I apply that university or private consultants look up table to that grid sample basically and so we've been researching this now for six years here in Kentucky and we found out that that's absolutely wrong. The recommendations have first of all extra fat built in and second they were built on trying to find the average rate phosphorus rate, and so we've got to talk about responsiveness versus total need and so on average if I take and and take all those grid samples in average and apply a flat rate to the whole field if that average soil test says I need phosphorus, we typically find that that's right. That it needs phosphorus, and I'll have an economic response across the whole field, but what we started doing was doing research where we took really small plots shotgunned across the entire field and what we did with those plots was we split them, so they're 20 foot wide, but we have 10 foot and 10 foot we're running four row corn planter and we're putting the phosphorus down in the two by two, and so we'll have a little 40 foot long stretch with phosphorous and right next to it. There's no phosphorous, so they're paired up so in that 20 foot by 40 foot space I know how many bushels I got by adding phosphorous because I have a zero phosphorus and a phosphorus at it. What we found is that fifty percent of the time in single digit phosphorus we use melec three, but you know really low soil test phosphorus 50% of the time I don't need any phosphorus. I can achieve max yield with no phosphorus even though my soil test is extremely low in that same field. When I take all those plots and add them up on average I need it faster and people go that makes no sense but when I respond, I get enough bushels to pay for the plots that didn't respond. But, it is random which plots respond when I go to the sub field level, so on average that fuel is low soil test and half of it needs phosphorous, but it's all low it doesn't matter what the soil test is. We tried everything we thought maybe it's the organic matter maybe you know maybe there's bugs living over here that aren't living over there and that has something to do with it. I think it's root length with phosphorus, so I'm gonna get off track for a second I think it's root length and phosphorus and this plays into the nitrogen story as well, but we need something else besides soil tests if we're gonna do precision ag and so this is phosphorus right. So, that so now we're starting to talk about this concept of response now in nitrogen when I started at Maryland we have this yield-based approach, so we thought about the same kind of thing like I've got a yield map, and I've got a recommendation, it says one pound per bushel, so let's take our yield maps and multiply it by one and that's the nitrogen rate for each of those zones. Let's zone it out let's do it let's say low, medium and high yield and unless it didn't work it failed 100% of the time it failed so using our yield map and using our yield based recommendation it failed. It was the wrong rate when I use the yield based recommendation and apply it to the whole field it worked. That was the right rate for the whole field and we can do strip trials with different rates and we can demonstrate that across the whole field it's the same thing as with the phosphorus that you have this kind of averaging effect and our recommendations are really accurate. So, on average they work but if we try and shrink the amount of area we apply that recommendation to, they fail because they're not precise enough and this gets back to the foundational work that underlies sensor technology like Green Seeker. So, Green Seeker algorithms are based on this concept of response and yield potential and they're independent and so that's hard for people to think about but they can move in different directions. You can have a low yielding field that's really responsive or you can have a high yielding field it's really responsive or vice versa, and so that algorithm is a traditional Oklahoma state model and what it says is at every second on the go sensor management at every second that Green Seeker is sensing the corn measuring NDVI and in the field we have a strip that before planting we put extra nitrogen on and another strip that we put no nitrogen on. And so the ratio of those two let's say the NDVI of that no nitrogen strip, I'm going into side dress at v6 and let's say it measures you know 0.4 and the one with high nitrogen measures point six right and so the ratio of those two point six divided by point four I should have used easier math is what things 1.5 yeah. All right there we go that's pretty good for a soil side just to do that on this problem, so it's 1.5 right it's the response index because that was pre-plant nitrogen says this is how much it responded on average and then as you go through at v6 and you've got your sensors on it's taking a reading every second of NDVI and it's predicting two things. The yield if I fertilize and the yield if I don't fertilize it's making that prediction it's taking the difference so let's say it predicts 200 bushels and it predicts 150 is the unfertilized yield, it says there's a 50 bushel difference and then it multiplies it by a factor, how much nitrogen do I need for 50 bushels and that's how it varies the rate, and so as you go across it's making this decision every second of what's my unified fertilizer, what's the unified what's that difference and so what you end up with is the really low NDVI, so that's poor small lots of soil showing through. The yield prediction is really low and the yield without is really low and so the difference is pretty small so the nitrogen rate is small right because it's a mass, actual mass of grain times mass of nitrogen. So, even if it's the yield is going to double it's still a smaller mass than in the higher yielding area, so just mathematically it's a low nitrogen rate. Then I get up into pretty good ground, and I've got a pretty good yield potential it's pretty much almost the maximum yield potential there that 50 percent responsiveness. Now, I'm talking about 200 bushels versus 100 bushels instead of you know 70 bushels versus 35 bushels. Right, so now all of a sudden I've got a bigger number to multiply times my nitrogen factor and my nitrogen rate's going to be really high in this good ground that's just below the maximum and then in the maximum we say well yield plateaus at some point. I can't get past this maximum yield and for this field maybe it's 350 maybe it's 400. 20 of the ground in this field yields 400 and this field is a 200 bushel field on average, and so I put in that max cap number so now you'll predict it with nitrogen taps out at that whatever I say is the maximum realistic yield, and when I hit that maximum realistic yield because that's a fixed number, but the yield without nitrogen keeps going up my nitrogen recommendation starts going down because what's happening is I'm getting into ground that if I don't fertilize we'll yield as well as if fertile growth. I'm going gonna grow 350 bushel corn here with or without nitrogen because it's just good ground and what we found was this algorithm worked like it failed sometimes mostly where it was tough to use was when there was a lots of residual nitrogen so we have a ton of chickens on the Delmarva peninsula right. So, we got historic year after year poultry application, poultry litter application, so you build up this big nitrogen like residual pool yeah, and so you're not responsive and so it's saying you don't need any nitrogen at side dress and it's usually right, but that's a hard pill to swallow, and so that's a problem and so there it says look you just need and then there's some other fields every once in a while the weather or something because it's predicting yield. I mean you're in there at v6 you don't know what the weather's gonna do for the rest of the season, so if the rain turns off it's gonna fall on its face, but guess what traditional management's gonna fall in its face if it stops raining after v62 we're all screwed. It worked really well and so on average Virginia tech reported that they expected a 20% reduction in nitrogen use and same yield statistically the same yield in corn and what we found was about 25% on average across our field sites reduction in nitrogen use in the same yield. So, now all of a sudden I mean it's paying for itself really quick, if you're cutting your total electric bill by 25% every year right it starts to pay for itself. I mean and so that I only work with the Green Seeker sensor and there's two other sensors on the market but Green Seeker specifically because of Oklahoma state's involvement in its development, it's open for algorithms that are local that can work not as a lookup table but as actually an equation the way it's coded in there it's going to read NDVI. The user puts in these these fixed kind of factors of maximum realistic yield, how much nitrogen have I already applied to date, and it's and the university can provide there's some coefficients in this empirical model that predicts yield and those have to be local right the other sensor packages don't allow that. So, I haven't worked with and they use a whole different philosophy they use a sufficiency. So, Nebraska kind of where optics came from right right exactly right so that was Kyle Holland's sensor, and it's based on this sufficiency approach which is like using a sensor to do traditional nitrogen management. I won't say it doesn't work, but I don't think it moves us in the right direction of more mechanistic nitrogen management. It's basically just another tool to do traditional nitrogen management if that makes sense. I'd be happy to try it out and test it and run them head to head but you know we just there's only so much we can do. So, we've always focused on this mechanistic approach.
Sam: You mentioned in some of the results of these trials that you looked at yield and nitrogen applied and what you're saving with that what environmental measurements are you taking as well to make sure that you are actually reducing these losses that the whole objective of this research is for?
Josh: Yeah, I mean so you hit the nail on the head, and I kind of have to whisper this. When we were working in Maryland the Maryland grain producers utilization board so that's like the corn, wheat check off, grain check off they were our biggest supporters. So, they really invested heavily in all this research and now here in Kentucky the corn board, Kentucky corn growers association, Illinois fertilizer checkoff. So, it's really farmers paying for this research right, and so I always kind of wondered out loud and some of the farmers that were friends of mine on the grain board back in Maryland said don't wander that too much out loud because you know you think about Green Seeker, and it's like it's varying the rate applying what you need where you need it. But, there are spots in the field where we need more nitrogen because it's leaky right and loss may be proposed proportional to rate apply I don't know right and like so most of my transport work was all in phosphorus. I've done some nitrogen leaching and some stuff in my career but most of my work was in phosphorus transport, but if we think about nitrogen loss I might need more nitrogen right here in this spot and so Green Seeker doubles the rate right here because it's basically a hole in the ground and all that nitrogen is going straight to the ground water or maybe it's that wet spot in the field where it's all denitrifying and that could be a problem. So, I may be cutting my total nitrogen use 20% and that has environmental benefit because just manufacturing transport of nitrogen is you know an energy intensive process right and so anything we can do to reduce energy uses has environmental benefit, so if I'm cutting my nitrogen rate 20% right off the top I've had environmental benefit for air quality and and those sorts of things, but if I look at water quality I don't know, and I guess I was always focused on the agronomics. Can I get the same yield and use 20% less nitrogen, let's hope at the end of the day the underlying cause of the Chesapeake bay issues is nutrient surpluses in other words I'm applying more than is getting to the crop, and we know that's an issue with nitrogen. So, let's hope that we're doing the right thing by it can't be that much worse than applying 20 percent more nitrogen yeah absolutely like on average we did a bunch of nutrient balance studies on the Delmarva peninsula and on average, we're applying 30 to 40 pounds per acre more nitrogen than we're removing with the crop on the Delmarva peninsula under best case scenario following university recommendations. So, that's if everyone's following the law and doing the very best job we think we can because nitrogen is so leaky we've got to over apply 30-40 pounds per acre right and that was at the end we did we did an evaluation of nutrient balances starting in 1998 going to 2008 over that 10-year period we were trying to figure out okay the law came into effect in 99 how much change occurred in this 10-year period and we reduced nutrient nitrogen surpluses, oh I forget what it was 60-70 percent, so we started out at like 70 pounds surplus and got to like 30 pound surplus. So, we did really good so we're doing way better and then if I add Green Seeker on I'm going to get rid of that 30 pound surplus yeah, hell I'm going to get rid of more than that right, and so now I'm in my balance and maybe then we need to start thinking about environmental, so now we need to start thinking about conservation practices and nutrient loss in a site-specific way because almost all the way we look at nutrient loss is using models that are at the edge of field. They're not site specific so now instead precision ag we need to think about precision environmental management and that's a whole new realm and that's so just recently I'd say in the last year or two I've been talking to a lot of other scientists around the country, at USDA ARS and some other universities, and so like I'm working with Amir Sadr report at southern Illinois university on this study and so starting this year we're going to be looking at denitrification with variable rate nitrogen from a sensor-based algorithm to say it you know again it's not necessarily looking at leaching, but we're going to be looking at denitrification and nitrogen loss and some of these environmental indicators and we're moving in that direction of okay now we understand the agronomics not to say that anyone's using Green Seeker. That's a separate conversation we can get to that in a minute but but if they were all using it are we cutting nitrogen usage 20% but increasing nitrate leaching because nitrate leaching all comes from this one spot in the field, and the rest of the field doesn't matter so I'm cutting all this nitrogen all over the field, but I'm pouring it into the one spot where it acts the one hot spot where I shouldn't be pouring it on this green seed or pour it on there so that's the next step is Green Seeker good or bad for the environment. My hypothesis is it's good, I think there's a better chance it's good than that right, but we don't know yet sure I mean as scientists we kind of have to be completely unemotional about this. I mean like the person Josh McGrath wants this workout right like I want to have a win-win, I want to see farmers have a lower fertilizer bill and the same or higher grain yields right like that's my personal interest, but my personal interest should have no bearing on how I conduct research and at the end of the day when I left Maryland one of the farmers on the green board I told him I was leaving Maryland. He said josh I never wanted to hear what you had to say, but I always had to hear what you had to say, and I guess I hope if there's any young scientists out there listening that that's I hope that I've lived up to that kind of expectation he had of me and I think that's what we need to do is just tell them the truth. They can deal with the consequences, maybe there's a law passed that says they can't apply nitrogen and they lose their farm right, but at the end of the day the first thing we need is the truth. And then the politics is separate like politics shouldn't always follow science. We think it should, but there are other human factors and economic factors all these things are real reasons like economics is a real reason not to do variable rate nitrogen, you know cropping up the Chesapeake bay is a real reason not to have the best nitrogen management practice, like if the very best nitrogen management practice is bad for water quality then maybe we can't have the best nitrogen cost benefit right, and so there are these other factors that society has to weigh in, but I'm the soil fertility guy like I can't have any emotional connection to my my research right you know and like my wife will tell you that that's my problem is like I'm incapable of emotional connection and but yeah that's right. But, it works out well for the job we've chosen right and like so we just need that data like we need to know.
Jack: So, you mentioned in there you said that nobody's really using the Green Seeker, and so it's my understanding a lot of these trials you did out there you know in the Chesapeake bay region were with farmers and you may have you know one or two that are still using them can you talk about kind of the farmers response to using the system what the logistics were what their challenges were and just kind of go into that that side of things as well?
Josh: Yeah, and this is a conversation that we just keep having and we haven't had we need someone besides engineers and soil scientists in this in this discussion, and I think it's you know we need that psychiatrist or psychologist or social scientist to understand human behavior right and maybe it's the software right. Like the software designer and the interface like that's the missing component to this, but generally when we look at the numbers we have a bulletproof way to do nitrogen management in the yield prediction and yield response equation of Green Seeker the body of scientific literature says this is this is correct in the right way agronomically to vary nitrogen in space and time from year to year and across the field in a given year use the sensor it is right and zero percent adoption nitrogen we know how to do it no one's doing it in Maryland. As I was leaving and part of this I think you know not that I'm the end-all be-all, but you kind of got to have someone driving these things and we were driving at the university and no one has kind of really filled in with that since we left but we had a program with a couple different non-profits and the one thing I can say is the policy is pretty onerous in Maryland on farmers and in the Chesapeake bay region as a whole but specifically in Maryland but in Maryland they were willing to pay like they pay a lot for cover crops as a result. 80% of all acres have cover crops, you can make more off of your I mean so everyone has a right well weak cover crop, so you're allowed to pay that's the other interesting thing. So, folks choose wheat over rte and get paid a whole lot more to plant rye, but they they'd rather have a wheat cover crop, so they'll take a lower payment but that payment is going to cover the cost of putting that in, and so they do it so we have 80% acreage coverage or better probably at this point but was 80 percent of all the row crop acres has got a cover crop every winter. You know, I'd be surprised if it's one percent anywhere else in the United States. I'd be shocked if you told me that five percent of Nebraska had a cover crop in the winter, so they're willing to pay so there was cost share programs that bought a bunch of Green Seekers and paid for some third-party support so anyway there was there was a bunch of farmers that got Green Seekers through some of these cost share and grant programs, and I don't have a good handle on all of them of what happened to them. I know some of the farmers that still use them because they still call me every year and send me all their data, and I look at it for them, they're still putting in field length, flat rate strips to compare to that Green Seeker which is smart right like because it's still a new technology you still kind of need to dial in some of your settings right and when they send that to me it's still looking good, it's still working out. But, my understanding is that almost all of those Green Seeker systems have been pulled off the rigs and put on the shelf in the shop since they don't have the support, so I don't normally talk about companies, but there's no way to talk about this without talking about companies and there has been no support of this technology now there might be product support from an equipment standpoint, but this technology like I don't care whether you use a Holland scientific sensor a top con sensor a sensor mounted on a drone or a Green Seeker right the equation is the equation I'm talking NDVI and I'm predicting yield with and without nitrogen. That algorithm is universal, if I have comparable NDVI and so it's an agronomic support that's necessary, and if we look at agronomists outside of an independent agronomic consultant who's selling nothing except for advice which is rare particularly on the east coast. There's no way to price in Green Seeker right because that farmer if he's buying fertilizer from you expects a rate recommendation as part of the package, and so do you charge per acre for using Green Seeker when you don't know how much nitrogen you're going to use till the end of the day. So, you say you pull up to farmer John's farm and you say okay I'm going to charge you five dollars an acre to use Green Seeker right. First of all, and you're assuming the farmer believes you they're not going to lose yield mm-hmm. I'm going to charge you five dollars an acre for Green Seeker okay how much my fertilizer cost me well I'll tell you when I'm done more or less than what you expect right. So, it doesn't work I mean there's a lot of reasons why this doesn't work. It doesn't work in that so first of all the farmer has to be doing all their underground and they got to put those sensors on and they got to be pre-contracting their UAM. So, now they're at that position now they gotta trust this black box they've got to understand the nuts and bolts of I'm gonna drive across the field and this magic sensor is gonna change the rate every second, and it's not like on the fly I'm gonna be grabbing the knob and turning it up or turning down if I don't like what I see you know it's like it's like hopping in you know a Camaro hashing the smashing the accelerator to the ground and letting go of the steering wheel you got no, you're you're along for the ride you've got no steering wheel you just take the wheel at that point, so psychologically this is an impossible technology to adopt without some support from the company that's selling it. There is none right I mean they might support the software, and they might support the hardware, but they don't support the agronomy right. So, you got to have the university I think the university extension has to find a new role for itself as that independent support if we think this is the right way forward right, but that's a big ask I mean you think about what county agents are doing already across the state and everything they've already got to do, and now they've got to be on call as tech support. It's almost a new role I mean how do we do that, how do we fund that how do we justify that, so this gets to something I'm excited about that's coming out is so farmers it's real on the go nitrogen management is right there's no question agronomically it's right, but it is impossible psychologically and I've come around to this you know I've been at this for a while and a lot of us sit around and talk about this. I think it's going to be impossible for most farmers to one in a hundred can do it and they're still farmers with Green Seekers doing the thing right. But, I think the majority of them psychologically won't be able to do this and so there's been some papers that have come out of Oklahoma state and Brian are now, I don't know if you all have talked to him yet, y'all might want to talk to him about this so they're saying okay well I can take a satellite imagery pretty high resolution NDVI and that NDVI like if it says this spot is point three Green Seeker is gonna say it's something else, but if you go out in that field close after that satellite image is captured with a handheld Green Seeker and boom boom boom, get some points I can come up with an equation that converts satellite basis that sensor basis, NDVI so now I take that satellite point three and I run it through the equation and it comes out Green Seeker point four is what Green Seeker would read there, it's not perfect but now we're we're getting into that gray area where Green Seeker on the go is perfect flat rate is what we do today let's get it something in between because the farmer kind of wants to look at that recommendation ahead of time. That's right you know chew on it a little bit, turn the paper around sideways turn it back upright crumple it up fold it unfold it and then at the end of the day it'll do what's on that paper that's how we treat fertilizer recommendations, and so I could take this satellite imagery convert it over to Green Seeker basis and apply the exact same algorithm to it and load it up as a prescription. So, now it's not on the go I know this is my prescription, and I'm going to apply a thousand gallons today whereas with Green Seeker on the go I may apply 500 gallons, I may apply 2,000 gallons right, and so I think that we're going to have to meet people where they are and I think sometimes in research we don't do that and we need to understand these psychological issues. It's not all soils and crops and rain and atmosphere, there's people in this equation and I think this approach coming out of Brian's program meets people where they are, so I'm interested in seeing how it goes.
Sam: Yeah, farming is definitely the ultimate risk, but when it comes to technology we farmers don't want to take that risk on that on the growth sensing.
Josh: So, well and look we kind of kid ourselves as farmers. If I can put on my farmer hat for a second that we're agronomists and soil scientists and plant scientists we're not we're like logistics is the business of a farm, how do I get this many how do I get this combined there how many bags of seed or bins of seed. How do I move this person there, how do I get my sprayer there that's 90%. What am I contracting my grain at, what did I buy my, did I buy 200 UAM last year or am I paying 500 UAN right now. Right, like that's it's it's a logistics business, so then to ask them to pick up this Green Seeker and figure this out on their own. I think it's too big of an ask.
If you'd asked me a year ago, I would have told you a totally different story I would have been like damn it everyone should be doing this. There's no reason not to, but I mean you know we evolve I mean I've spent a lot of time all the research we do is on farm. We don't do anything on the university, so there's a farmer involved in every set of plots I do. That's you know so I spent as many hours in the combine as I do talking to farmers and so over this time you just kind of feel your way out and kind of feel out these other factors and you say we got to come up with something different.

Sam: Yeah, I just defended my thesis on a lot of the things that you have covered, so you're definitely.
Josh: Do you agree with me or disagree.
Sam: I very much agreed with you on basically all of it. I have a bajillion things to add, but I think where we need to go especially to be courteous of your time so where do you see us in the next five years where do you think we should go to overcome some of these challenges?
Josh: Well, in the next five years I was hoping you would take a longer time period so I could just say it's your job
stretch out a little bit further I won't be involved in this question anymore. I think there's multiple things like there's a kind of engineering if a bridge fails it's never because of one thing there's always like 20 things and the same thing with the success right. I think there's a lot of difference, so I think we need to reach out. We need to get some people that understand the human how do humans interact with data so set aside Green Seeker and grid sampling all that. I mean farming precision ag is about it's not even precision ag anymore it's about you know data based agriculture we have at our disposal just you know gigs and gigs and gigs of data. You know second base weather data there's going to be bi plant sensors that dissolve in the field that you that are inserted when the corn is planted. I mean back in the early 90s the original Oklahoma state work figured out that the right space to be varying nitrogen was on a on a plant by plant or two plant basis every seven inches we should be changing the nitrogen rate and from a machinery standpoint that's actually really easy to get to like if we think about processing power data storage and machine control we could probably do that today if it wasn't a factor of cost. The only thing left is for the price of those things to come down and we could be putting flocks of robots out that you know have micro pumps that are shooting nitrogen every plant this is easy it's the agronomy and it's how the people intersect with that data, and so I think we've got to start training scientists that are experts on how humans interact with data interpret data and use data and that's different than anything I've ever faced in my career. I mean there's whole disciplines that are about how we internalize data in our heads and how we how we plug into that data you know and you know sci-fi movies like the matrix got into this like how how do we interact with that data and so I think we're on the precipice of that. I think that's where things are moving is data-based farming and understanding how people interact with the information and designing tools that facilitate that interaction. So, can you design all right we get a halfway you know satellite based prescription map versus reactive approach to get people's toes into Green Seeker right. But, five years from now we should still be trying to go to on the go every plant gets an individual rate nitrogen management because we know that's better and so that's going to require designing that technology to interface with the person in a specific way that they are comfortable with it right, and I don't know what that is and I will never know. I'm not smart enough to figure that out, but we need to get some smarter people involved in this you know like you all to figure that out. I mean that's it's so I think that that's where we're going to go is that the interaction with the data is going to become very important.
Jack: Absolutely, it's a totally different challenge it's basically the first generation right that's ever had to deal with this fact that we have so much information all the time what do you do with it, how do you make sense of it and how does that challenge your preconceived notions. So, I don't know it is going to be a big challenge, and I think it's going to require as you mentioned an intersection of engineering and agronomy and the social sciences because that's really where it sits. It's kind of in that nexus, so I guess Josh to kind of finish this up what is one piece of advice I mean we've been talking about nitrogen management we've talked about phosphorus we've talked about all these different concepts that I think people probably don't think about at such a detailed level right. We talked about these sensors that nobody necessarily is adopting right now what is your piece of advice for somebody who's wanting to improve their precision nitrogen management?
Josh: Yeah, so I think at the end of the day where this whole conversation has taken me is the sensors attempted to look at the mechanisms of nitro requirement, so if we go backwards in time in 1965 George Stanford was working in Beltsville and Maryland at the ARS, and he came up with Stanford's equation right and that's the basis of how Green Seeker works, and it basically said there's there are mechanisms to nitrogen requirements like how much yield how much loss and how much nitrogen per bushel of grain but the total requirement to make the plant and the grain and we can figure these things out it's hard to predict on a day-to-day basis. Mostly that loss and the yield potential because of weather, but he said and so that paper no one ever talks about what I think is the most important part of that paper in 1965 at that time you either have mechanistic models or you have empirical models at that time everything was empirical which is what mrtn is. I'm going to go to 100 fields this year and we'll put out 20 nitrogen rates and one is going to optimize yield and then I'm going to take that average across all those fields and that averages 180 pounds of N per acre. So, on average in the state of Kentucky I need 180 pounds of N. That is an empirical recommendation and Stanford said we will never advance the precision of our nitrogen recommendation unless we start looking at the mechanisms of requirement and now we can split out like we can split out yield into different things. I mean you can start making that equation really complicated but my one piece of advice to the farmer or the agronomist or the CCA is to think about nitrogen mechanistically understand the nitrogen cycle and if you want to dial in your nitrogen management again timing is the big lever right like that's where I can get the most advantage for my my client or my farm is by adjusting the timing to match crop need because nitrogen is so mobile. So, we think about that mechanism of timing up until v6 a 200 bushel crop probably only needs about 20 pounds of nitrogen, so I only need 20 to get to v6. I don't know what's going to happen or what the weather's going to be, so I'm going to put 75 down to just make sure I get there right because it's cool in the spring there's not as much loss, so I'm starting to think about these mechanisms when it's warmer you know they're greater, faster mineralization more loss ammonia volatilization and someone says okay I need 20 pounds to get to v6. I'm gonna go 75 and now I'm going to think about what I've got left right, and that's where I'm going to start, and I'm going to say what was my crop rotation what does that do did I have a rye cover crop. Well, I got to overcome that carbon penalty, did I have soybeans before okay I can cut back a little bit because that residue. So my advice is one real simply think about timing forget wide drops they're great spend money on them if you want to if you're just saying how do I side dress take your sprayer get some nozzles talk to an engineer about nozzles put a piece of hose on there a piece of hydraulic hose slide a piece of PVC over it so it's not flopping all over the place and just drag a piece of hydraulic hose through the field and you know dribble on that side dress start at v4 when that corn's about 12 inches tall and hopefully you cover all your acres by v10 when you can't get through the corner without knocking the tops off right right I mean it's simple. I think sometimes we make things too complicated but that's that's my simple advice start split applying and then start thinking about the mechanism.
Sam: Thank you so much to Dr. Josh McGrath for joining us today on the FarmBits podcast, and it was so fun to talk about nitrogen management specifically sensor-based nitrogen management since that's what both of our research is in. I think my favorite part was talking about when it comes to nitrogen management or nutrient management in general about how sometimes it's these bigger ideas, and he mentioned timing a lot it's those bigger controls that mean a lot more than we're trying to fine-tune some things, and we have to be sure we're taking care of the big things first it almost makes you think about Terry Griffin right, if you can't get the regular stuff right precision ag is not going to do anything for me.
Jack: Exactly and I thought that was another thing that was really interesting is Josh kind of brought the idea that your user has to be comfortable and has to trust the system, and I think that's huge. I mean it's a really big challenge with proximal sensors and on-the-go sensing but just in general with any sort of technology out there you have to build that trust with your users that are going to be implementing it on the field. 100% so that was a really good episode really excited to talk about this type of technology and next week we're actually going to flip the script, sam and I are going to be on the other side of the mics. I hope you'll be around for next week as we dive into remote sensing and a little bit more on proximal sensing with nitrogen management. See you next week.