Field Crops Webinar Series - Tillage in Michigan Soybean Systems

February 25, 2019

This webinar features James DeDecker, director of MSU Upper Peninsula Research and Extension Center, and focus on tillage in Michigan soybean systems.  In 2016, a team of MSU Extension educators, faculty and growers set out to gather on-farm data, aiming to help Michigan soybean producers better target tillage to improve their yields, profitability and sustainability. Our study identified environments where no-till lives up to the hype, as well as areas where conservation or conventional tillage would be recommended to maximize soybean performance and profitability. In the latter case, we looked at the inherent capacity of soils and complementary practices including manure and cover crops to buffer soil disturbance and protect soil health.

Video Transcript

- And with that I'm going to introduce Doctor James DeDecker. He's the director of the MSU U.P. Research and Extension Center. And today he's gonna be talking about tillage in Michigan soybean systems. So James, I'm gonna stop sharing my screen here and if you're free to talk. - Okay, thanks Ricardo. Thanks for that nice introduction. I do have to give you just a little bit of a correction. My degree isn't quite minted yet, so can't call me doctor just yet and sometimes I feel like it's almost bad luck if we use that preemptively. So I'm just James for tonight. Thank you though. So I'll pull up my slides and get that all set here. All right, can you see that? - Yes we can, all good. - Okay. All right, thank you very much. Well, as you can see, my title tonight here is Targeting Tillage to Optimize Yield, Profitability and Sustainability in Michigan Soybeans. I'm gonna be talking about approaches that we took to understand the implications of tillage in soybeans here in Michigan, and information that we gathered to help growers understand what tillage approach might be best for their given system and location. Oh I should probably say, before I move on. This work was funded by a few different sources. Michigan Soybean Promotion Committee as well as Michigan State University Extension and MSU Project GREEEN. And I'll also give some acknowledgements to the team that was on this project before I jump in. I had a couple of colleagues in Extension, Marilyn Thelen and Eric Anderson. Eric is on the webinar tonight. As well as my adviser, Doctor Adam Davis and committee member, Doctor Sieg Snapp, at MSU. And a fellow student, Xinyi Tu, at MSU. So thank you everybody who helped out with this work. A quick overview of the talk tonight. I'm gonna start off pretty generally, with a definition of tillage and not that you don't know what tillage is already, but I wanna emphasize some points there. And then I'll talk a little bit about the problem that we're trying to address. The specific issue of deciding what type of tillage approach to use in soybean production, specifically here in Michigan. And then we'll talk a little bit about some of the objectives and questions that we had for this study, or approach, and then I'll spend the bulk of the time on our results. So what is tillage? At it's most basic definition, tillage is mechanical modification of soil structure. But what I really wanna emphasize here is the interaction between soil and tools. So tillage tools modify soil structure through this wide range of soil tool interactions. And the outcome of those interactions varies with respect to the characteristics of both the operation itself and the characteristics of the soil that's being tilled. And so trying to emphasize those interactions and the variability of those outcomes, because what really drove my initial interest in this topic of tillage in soybeans, is that you often hear kind of blanket recommendations that, you know, for soybeans in Michigan you ought to use no-till, because that's the best approach. Or maybe you ought to use conventional tillage 'cause of the kind of state that we're working in here or the production systems that we tend to have here. But what you find is that if you implement a single tillage system like no-till, across a wide range of situations and conditions, you end up with different results, right. So you really can't make black and white recommendations for much in agriculture at all, and that also applies to the topic of tillage. There's an issue in tillage research, of trying to quantify tillage intensity and really define tillage systems. So I wanna touch on that a little bit. One way that we can define tillage intensity is simply categorically. So I mention no-till, right. But a lot of people have different ideas about what exactly that means. Does that mean I never till the soil? Does that mean the only disturbance that I have is that drill going through the field at planting time? And you can also talk about something like conservation tillage. What does that mean? If you're NRCS, that used to mean 30% residue cover remaining after planting. So after all your tillage operations, your planting operations, you had at least 30% residue cover and there's an image from our project about what 30% residue cover looks like. There's also another way to categorize or quantify tillage, and that's with this approach called STIR. It's a Soil Tillage Intensity Rating, and that's from an NRCS formula, the RUSLE2 model, that helps quantify erosion, or potential for erosion. And so with the STIR model, you can see that there's different tillage operations in this table here. And they each have a STIR value or a STIR coefficient. And so basically, depending on what tool you're using, how it's set up, each tool has a different rating in terms of how intensively it's disturbing the soil. So no tillage is a zero. You move towards the middle here, vertical tillage might be something like a 20. Moldboard plow would be a 55 to 65. And those STIR values are on a range from one to 200 overall. So there's different ways that we can think about that. We can also match up the STIR value to these categorical approaches, to kind of quantify what tends to be in a more categorical system to defining tillage. So when we think about how growers are making tillage decisions, there's a lot of motivations to do some tillage, and there's also a lot of good reasons to maybe thinking about reducing disturbance in your system. So motivations to till. A lot of things like seedbed preparation, controlling weeds or pests. Trying to limit pesticide input. So maybe I can control weeds with cultivation instead of herbicide, for example. Incorporating amendments, fertilizer. Residue management can be an issue sometimes and a reason why we use tillage. There's also some things that are a little bit more nuanced, like aesthetics or tradition. Maybe I've always used a certain type of tillage system on my farm, or maybe I like the look of a clean tilled field, and I'll get more into kind of the human dimensions, or the social science aspects of this topic later on in the talk. But those come into play as well. On the side of not to till, things like economics, right. One of the biggest motivations to reduce soil disturbance would be saving on things like equipment costs, fuel costs, labor costs. We also hear a lot about soil health, right. Can I reduce erosion or build soil health? Build things like organic matter or carbon concentration in my soils, to try to have a healthier, better functioning system. And there's also, again, kind of a social science concept that maybe reducing tillage or adopting no-till is the right thing to do, a more progressive choice or more innovative approach to farming. Is that the kind of, the next best thing that maybe a grower should consider. When we compare conventional tillage and no-till systems in research, it also kind of speaks to some of the motivations for why you might wanna reduce disturbance in your system. So these are a couple of radar plots, they're called and this is from work coming out of Kellogg Biological Station at Hickory Corners in Michigan. One of our research facilities here for MSU. And you can see that if we compare these conventional no-till systems on different parameters, for example, grain yield on the left side of these plots. In the case of KBS, actually no-till yielded a little bit more, on average, than the conventional tillage system. So hey, that's great. But we can also look at things like ecosystem services. Things that are important to sustainability like soil organic matter. And we can see that, in no-till system, they were able to build a little bit more soil organic matter, than in the conventional tillage system. That contributes to things like soil water holding capacity. So a lot of good reasons why people talk about reducing disturbance, reducing tillage, adopting no-till. But do those things always hold up? In soybean, specifically, if we look at, across the United States, the adoption of conservation tillage. Here in the blue we have mulch till and we have no-till in the yellow. And you can see that if we look across these common crops, wheat, corn, soybeans, cotton, actually adoption of conservation tillage and no-till in soybeans, is the highest of just about any of these kind of conventional field crops that we have in the U.S. So we've got, on average, 2006, 2012, right about 70% of planted acres for soybeans were in some sort of conservation tillage system. Either mulch till or no-till. So overall we're seeing high rates of adoption which would suggest that people are having success with conservation tillage in soybeans. We can also think about what does no-till or conservation tillage mean for yield. On the left hand side, this is from a global meta analysis that looked at the yield effect of no-till in different crop categories. And here, if we look at legumes, you can see the effect on yield was really negligible. So on a global basis, adopting no-till and soybeans had almost no negative effect on yield so, kind of like that KBS example, if we can have a win-win situation where I have competitive yields but also have all these environmental or conservation benefits, save on equipment, fuel, labor, why would I not adopt no-till, right. And if you look on the right hand side, this is an example, again, from KBS in Michigan where you can see even on a smaller scale, or more local example, the no-till system actually out-yielded the conventional system. So here on the Y axis, one is the average yield of a conventional system, and so if you're above one that system out-yielded the conventional system. So the biologically based, kind of an organic type system yielded less but reduced input and no-till were competitive yielders, or maybe even a little bit better. So what's the problem, right. Why did we wanna study this topic of tillage in soybean systems for Michigan? Well when you start to break it down a little bit further and look at the implications for tillage, adoption of different tillage systems on a regional basis, the story changes a little bit. So at the left here we've got the United States broken down into different regions. And what you can see in these circles, we've got overall adoption of no-till or strip till in the light color. Kind of the cream color here. And then we've got other tillage systems in the orange. So this is across all different crops, but I've also added some numbers here to show you adoption of no-till or conservation tillage in soybeans specifically. So the black numbers here on the left, we've got adoption of no-till and on the right we've got adoption of conservation tillage. And this, again, is specific to soybeans. So what you see here is that if you look in the southern areas, for example, a prairie gateway in the purple here, or if you look in the heartland areas, you can see that adoption of no-till and conservation tillage in soybeans is a bit higher than we have here in the northern crescent area. So our adoption of no-till for soybeans is 35, conservation tillage 26%, versus down in the south, they've got 71% no-till, 22% conservation tillage. Or in the southeast, 41% no-till, 22% conservation tillage. So my question is, why are we lagging in no-till adoption in the northern crescent, or in this kind of northern tier of states in the United States. We can also look at tillage practices across time. This is tillage systems on a national basis on the right here, across years 1996, 2002, 2006, 2012. And what we see that, on a national basis, we had an increase in no-till for the first three of those years, and then, when we moved into 2012, we saw a little bit of a drop in no-till adoption. And if you look over at conservation tillage, on the far right, basically as no-till is trending up, conventional tillage, I'm sorry, was trending down, and then we get to 2012, there was a little bit of a swap where we saw conventional tillage pop up and no tillage decrease a little bit. So what's happening there? Well a couple of different things you might consider. One thing it's possible that people that had adopted no-till found that it wasn't working as well as they wanted to and they reverted back to a conventional tillage system, or maybe a more middle ground conservation tillage system. What could be some of the failures there? Some things you could think about would be things maybe like herbicide resistant weeds. Maybe the yields in no-till were trending differently than they wanted them to. Or another thing could be something like oil prices. But basically there's some changes in tillage practices that we need to try to dig into and understand a little bit better, on a regional basis, and even across time. When we zoom in a little bit, or take another approach to try to look at these regional differences in tillage performance in the United States here, we can see kind of a latitude-based trend. So this is another meta-analysis from DeFelice et al, that looked at a different region in the United States, across a range of long-term tillage trials. What was the implication of these, of no-till on yield in soybean specifically? So what you saw here, it's kind of similar or might help explain some of the trends in adoption that we saw on the last slide. So in the southern tier here, in the green, we see that, overall, there's more blue squares than there are red triangles. And if you look at the table here, the southern area, this is the percent advantage of no-till. So we see positive numbers here for the south and the western region, and so that is suggesting that overall there's a yield benefit to no-till soybeans in the south. As we move our way north here to this, what they call the transition zone in blue, we can see that there are some blue squares, a lot of yellow circles, a few red triangles and if you look at the table here you can see that it's a little bit of a wash. There's a slight negative yield pattern with no-till, in that transition zone, but it's very close to zero. When we move up to the north, Michigan and Wisconsin, a little bit of Canada there, you start to see a lot more red triangles and if you look at the trend overall, we have negative numbers, suggesting there's actually a yield penalty for no-till in these northern tier areas. And if you look at these different conditions, they've got soil drainage and crop rotation. So basically, we can look at different types of soil. For example, on a poorly drained soil, the yield penalty for no-till is greater. 6.4% versus 2.4. And with crop rotation and continuous cropping systems, we've got 6.4 versus 3.6 in a rotated system. So, essentially, the further north you go, the worse no-till performs in terms of soybean yield, and that is accentuated by poorly drained soils or continuous cropping systems. So that led to the questions that we had for our study here in Michigan, and we wanted to understand a little bit more about how does tillage interact with the biophysical environment to affect soybean yield. And then, what about soil carbon because no-till is promoted for its soil health benefits. So we wanna understand that relationship with soil health in terms of soil carbon as well. And then the ultimate driver is really how can we make tillage recommendations that account for this environmental context, and grower decision-making behavior, as well, I mentioned some of the social science component, to try to balance a soybean yield and soil health. Some of the methods that we used. We tried to work with growers directly on their farms for this project, and we did that in terms of grower learning communities. We sampled commercial soybean fields and then we used what they call data-mining techniques to try to build some predictive models to better understand these relationships. - [Ricardo] Hey James! - Yeah, go ahead. - [Ricardo] Can I step up just one sec? - Yeah absolutely. - [Ricardo] We got a question here. I don't know if you have access, do you want me to read that out loud or you wanna? - Yeah, why don't you read it out loud. I can't seem to pull it up while I'm sharing this. - All right, sounds good. So John is asking that, and he's saying the reason that because this is because of no-till that the soil does not warm up as fast in the north to get the seeds growing as soon. He's asking that. Do you want me to read that again? - No no, that is a very accurate way of summarizing that. So, essentially, the big challenge with no-till is that you don't have that disturbance to warm up and dry out that soil in the spring. So we talk a lot about plant date in soybeans. There's a great advantage for planting earlier in soybeans, and so if you do that and you plant early sometimes you're planting into a cool soil, a wet soil, and that is where no-till really suffers. And so that's probably the biggest driver for that pattern, that regional pattern, as we move from south to north and why we see more struggles here in Michigan. So absolutely. And we'll see how, as we move forward in the talk, we'll see how that plays out in terms of soil conditions too, right. So if you're on a well drained soil versus a poorly drained soil, for example. So yeah, great question. - [Ricardo] Awesome, thanks so much. - Yeah. So we worked in these groups of growers that we call learning communities, and that was really to emphasize that we wanted to go through this project, side by side, with growers, to learn together. So they were learning from us, we were learning from them, by regularly interacting as we conducted this research. So we looked at 35 different growers in three parts of Michigan, and recruited them a couple of different ways through Extension, through the Soybean Promotion Committee, and we met with them about three times, annually, throughout the project, and so the main project occurred in 2016, 2017. We continued to work with them in 2018. And we're even doing some additional followup today. This is where our sites were located. So we kind of had a transect, if you will, across the lower peninsula, where most of the soybeans are produced. And we were sampling at 133 commercial soybean fields, and you can see the breakdown by the regions of the state there. And we sampled these by soil type within field. So we wanna understand the soil variability within fields. So that broke down into 273, what we called zones, and we used quadrats, or squares out in the field, if you will, flagged off areas, to sample the crop within those zones. We called this approach soybean epidemiology, and that is just a fancy way to say that this was an observational study. So we weren't asking the growers to change their practice in implementing treatments and experimental fashion. We were simply collecting data about the existing conditions out there in the field. Similar to someone, how they would approach understanding disease in epidemiology. So we had four fields per grower that they self-identified as good versus bad. And we had them maintain their practice, so you can see the breakdown of folks that we had in long-term no-till conservation and conventional tillage there, by percentage, in our sample. And we established these soil zones in the geo-referenced quadrats in their field. We visit each field four times a year for sampling. And on the right here are the things that we tracked in these fields. So we did a lot of soil monitoring. Both chemical, physical, biological parameters. We also did crop monitoring. Looked at stand establishment, did some tissue analysis. Quantified yield. We had growers fill out a pretty extensive survey about their management history on these fields so we can understand what they've done in the past before we got there. And also what they did in the years of the study. And then we did some social science too. We looked at human dimensions to try to understand how they decide what tillage practices to use. Oh that's a bit of a repeat. So the first thing that we did in our approach to analyzing this data is we wanted to group these soil zones or kind of a subfield level, based on environmental similarities. So we wanted to understand what environments were more similar and more different. And we did that by looking at location, weather and soil quality variables and we grouped or clustered these sites by their environmental characteristics. And so location and weather included their actual location, latitude and longitude. Precipitation. Reference ET which is a measure of evapotranspiration. Their growing degree day accumulation or temperature at certain thresholds. And then soil quality was texture, carbon, labile carbon which is that more active or rapidly turning over pool soil carbon. Cation exchange, soil pH and also penetration resistance. We had to choose how many groups to delineate and we used this approach within group similar squares. We came up with a six cluster solution. And you can see how they're grouped here by a couple of different principal components but probably what makes more sense is looking at this approach, clustering on the map. So here you have these different environments. We came up with a six cluster solution. So we identified six environments that within each of these clusters, there was a lot of similarity and then between clusters there was as much difference as we could manage. So A through F here, and you can see this shapes and the colors that identify those. We essentially had two clusters in the northeast, A and B. We had C and F primarily in the central part of the state and then cluster E down in the southwest. Cluster D was a bit of an outlier. It only included one case and, if you look at, particularly the organic matter numbers here, 8.5 average soil organic matter. That was a real outlier, a muck soil. We ended up throwing out cluster D. So in the end we had five different environments, or clusters, A, B, C, E and F. If you look at the soil characteristics you can see that in the northeast we had kind of a coarse-textured cluster B with a lot of sand, and a finer textured cluster A, with a lot of clay. And you can see the organic matter and the labile carbon and the POXC numbers for those. In the central part of the state we had, again, C and F and sort of a separation between the finer textured soils in F versus the coarser textured soils in C. And then cluster E, down at the southwest, was a quite coarse-textured, sandy soils. Very low organic matter. 1.3% on average. So if you've farmed in any of these areas of the state or you travel in any of these areas of state and looked at their soils, their production systems, you can get a sense of how this kind of breaks down relative to those patterns that you see out in the field. Next thing we did is something called yield gap analysis and basically wanted to understand what's the difference between the yields that we measured in the field and the yield potential, based on these environmental variables. So the yield gap is the maximum observed yield in that cluster, in each of these environments, minus the zone yield. So what's the difference, again, between the highest yield that we saw, kind of that observed yield potential, versus what each grower actually achieved in each of these sites. Then we used something called linear mixed effect modeling to try to determine the effect of tillage intensity on yeld gap, and we wanted, again, to kind of understand if you can choose a tillage practice which one is gonna help maximize your yield or reduce yield gap. Get you closer to that yield potential. And this is kind of a breakdown of the result here. So you can see, along the X axis, the different clusters or environments that we had. A, B, C, E and F. And then the yield gap on the Y axis. And, again, remember this is the yield gap so you have to kind of flip the way you think about yield. We're trying to minimize the yield gap or maximize our yield to get closer to that yield potential. So what we saw is that different environments require different tillage systems to do that and I'll start with where no-till was successful. So no-till is the blue bar here and you can see in clusters C and E, no-till was competitive. There was no significant difference in yield gap between these tillage systems. And so the recommended system was no-till because there could be some environmental benefits. There's certainly gonna be economic benefits. And there's no yield drag. So it would be a great approach to use no-till in environments C and E. However, if we look at A, B and F, the story changes a little bit. In A and B, remember these are our environments in the northeast lower peninsula. The recommendation was actually conservation tillage in both these systems. In A, conservation tillage actually reduced the yield gap quite a bit. So we had better yields in a conservation tillage system. And similarly, in B, actually, in environment B, the conventional tillage system further reduced the yield gap, but the other thing that we took into account in these recommendations, was the tillage cost. And I'll talk about that in a second. If we move over to F here, just looking at the influence of tillage system on yield gap. Conventional tillage was actually a benefit in F, relative to no-till or conservation tillage. So the recommendations here, as I mentioned also, take into account the cost of tillage, so it's not just the influence on yield gap but what is the cost of tillage, and we took this from survey data from the firm team here in Michigan, our business management group. They surveyed growers and asked what they pay for custom management practices. And average cost of tillage per acre per pass is 10 to $25. So you gotta account for not only the yield that you're gaining by implementing, say, conventional tillage in cluster F but what is the cost of doing that tillage and still, because of the dramatic reduction in yield gap, tillage not only improved yield but actually paid for itself as well. So here's an economic approach to say that in cluster F conventional tillage was the recommended system. If you recall the differences in these environments, again, A and B were in the northeast, so they were our furthest north or, you know, the question that was raised earlier about temperature and moisture in the spring. These are gonna be our coldest environments in this study in Michigan, and so it would make sense that you maybe need more intensive tillage in that colder environment to maximize yield potential. And if you remember cluster F, that was in the central part of the state, but it was some of our heaviest and, I shouldn't use that term so loosely, but this was some of our finest-textured and highest organic matter soils in the study. And so similar to that meta-analysis we saw at the beginning, on higher organic matter ground, higher water holding capacity, poor drainage, you're gonna have more of a benefit to tillage. - Hey James! - Yes. - [Ricardo] Okay, it's me again here. So I think you kind of answered the question but I have a new question here. But maybe you can explain to John a little bit better. - Sure. - So he's asking here, any reason why there is no test plots in high production areas like the Thumb or Saginaw Valley. Do you expect different results if those areas were included? - Mm hmm, yeah. So it's a really good question. Our sample is, in some ways, a convenience sample and so there are questions about how would other environments play out in a similar study. So if you remember, we recruited our growers through Extension. I had colleagues, Marilyn and Eric, in these different parts of the state and so we went to where we had access to growers and access to their fields. So it's certainly possible that if you went to other parts of the state you might get different answers. But I would suggest that the general pattern of as you move further north in the state, there's gonna be more benefit to tillage, and as you get on soils that tend to stay wetter or have higher organic matter levels, that you're gonna see more benefit to tillage. So if you're on some muck soils in the Thumb, old peat soils, maybe you would see a benefit to tillage. There's also factors like artificial drainage, right. Out on the Thumb a lot of those old lake-bed soils are artificially drained. So although they're higher in organic matter, maybe they still dry out relatively quickly. So yes, all of those things come into play and there's a great need when we're doing this kind of observational work, to go out and verify it. A goal that we have for the future is to actually implement tillage treatments in different environments, to see how that plays out. But yeah, it's a really good question. It'd be nice to do this work elsewhere as well. - [Ricardo] Awesome, thanks so much. - This is a lot of numbers and figures here. I don't expect you to grasp all of this but I did want to raise this point. Some additional analysis that we've done looks at the question, what if my tillage system is fixed. So in the previous approach, we were really keeping the door wide open in terms of what tillage approach you could use. Now this analysis was asking what if I'm gonna no-till. So say I bought all the no-till equipment, I converted my system. I'm set on that. I'm not gonna implement conventional tillage. Or maybe I'm really sold on the soil health benefits or the economic benefits of no-till and I wanna try to maximize that. How do I maybe select a field or adjust my other management practices to try to optimize that no-till system? So basically, what this is showing, is that, again, things like soil texture in this second column here labeled PC2, present sand, silt and clay, and organic matter and its relationship with cation exchange capacity, on PC1 here on the left column. All of these things are important for determining the performance of no-till, which is good. That aligns with what we talked about earlier. And also we looked here at the interaction with plant date. And this was pretty interesting. So here, in these various figures you see on the Y axis arranged from zero to one. And zero is low probability or no probability of being a high yielding no-till system. And one is absolute probability. 100% you're gonna be a high-performing no-tiller. And what we see here is interesting, that there is a relationship with organic matter and texture, that, if I'm on a high organic matter soil, the probability of being in a successful no-till system, a high-yielding no-till system, is lower, and that's PC1. But as we move forward in plant date, so these individual boxes are labeled plant date day of year. So this is the calendar day of year. As we move from 134 to 136 to 138 to 140, what we see is that the probability regardless of soil texture, the probability of being a high-yielding no-till system, gets better. So when you go all the way to plant date 158, you can see that we're able to compensate for some of that risk of no-till by planting later. Now there's some question about well, in general, I've heard that planting soybeans early is gonna give me higher yield potential. So you could look at this a couple of different ways. If you're dedicated to a no-till system, delaying planting until the soil warms and dries is gonna be an advantage for you. You won't see as much of a yield drag. Certainly you don't wanna wait any longer than you have to. The other thing is, if planting is delayed and you have a choice of tillage system, you might consider putting in your no-till or putting in your late plant of beans with no-till instead of conventional or conservation tillage, because you know that you're not gonna see as much yield drag from no-till. So we can adjust other management practices to try to alleviate some of the issues that come along with our chosen tillage system, or some of the environmental conditions in our fields. We also wanted to understand the relationship with soil carbon, and so we looked at the relationship between tillage intensity here. Variable two is STIR, remember that Soil Tillage Intensity Rating and POXC. POXC is permanganate oxidizable carbon. That is a measure of labile carbon or carbon that cycles quickly in our field. And what we saw is that in two of our clusters, over here on the left side, A and B, remember those are the clusters up in the northeast, there was a relationship between the level of tillage intensity and the level of labile carbon. And so this is a hint for us to suggest that there is some relationship between tillage intensity and labile carbon, as contributing to the difference in yield gap that we're seeing. And one thing that has come up in looking into this further is that as you move further north, and temperatures are colder, we tend to have more carbon accumulation in soils. In particularly that labile carbon, that normally in a warm condition would cycle quite rapidly, doesn't cycle quite as quickly. Think about, if I heard about issues with the permafrost melting in the Arctic and how there's a bunch of carbon locked up in that permafrost, and if that melts we're gonna have a lot more carbon going in the atmosphere. If you think about that at a lesser scale, even in Northern Michigan, there's more carbon and labile carbon that is locked up and not cycling as quickly, because of lower temperatures. And so by increasing tillage intensity in these northern systems, in these higher carbon situations, we can cycle some of that labile carbon and get the benefit of cycling that carbon in terms of nutrient mineralization and biological activity that helps make other nutrients available to the crop. And so we often talk about building soil carbon like we wanna just fill that bank account, right. We wanna get as much carbon as we can in the soil. But really, it's not necessarily just having carbon in the soil. I mean there is some benefit from that. Water-holding capacity, for example. The more carbon you have, the more water you can hold in your soil. But some of that benefit from carbon actually comes from cycling that carbon, from burning off CO2, unfortunately, by doing something like tillage. Increasing biological activity, increasing temperature. Drying soils that may otherwise be anerobic. And getting that system moving. My adviser calls it the smokestack model. We need smoke coming out of that stack for the factory to be running. We wanted to understand now if we implement tillage in some of these conditions, to try to maximize soybean yield, what's that gonna mean for soil carbon. And one thing we looked at is that labile carbon by tillage recommendation domain. So in those zones where tillage is not recommended on the left, versus where tillage is recommended, we can see some general differences. One is that overall we have more labile carbon in the systems where tillage is recommended, and also we tend to see more of a pattern between labile carbon and tillage intensity, where tillage is recommended, versus where it's not. So, again, we seem to be kind of cycling that labile carbon with tillage. And then we also looked at total carbon because the risk is, if I till to maximize soybean yield and burn off all that labile carbon, am I going to deplete my soil carbon bank and end up in a situation where I've got degraded soil and maybe the benefit of tillage disappears and I have erosion and all the problems that come with too much soil disturbance. And, in general, actually that's not really what we observed, at least in this dataset, that it didn't seem that we were losing soil carbon overall, total carbon, by tilling more intensively in these areas where tillage was recommended to improve soybean yield. So that was kind of hopeful. Another thing we looked at, kind of generally is, what sort of things can we do to offset or buffer the soil disturbance that we are recommending in some of these environments. And one thing was manure. We asked growers, do you use manure or no. Now this is a very general comparison. We're not asking them what type of manure, how much manure. We just asked them yes or no. And what you can see is that really there's a difference between systems where there's no manure and there was manure when you look at the relationship between tillage intensity and active or labile carbon. And basically, if I'm not adding manure, as I till more intensively, I am having less labile carbon. However, if I'm adding manure actually I can hold off that reduction or even maybe see an increase in labile carbon. If you think about it, it's kind of carbon in carbon out. So yes, I'm tilling more intensively but I'm also adding a bunch of carbon, in the form of manure, and so that could be a net benefit to tillage if I'm incorporating manure with that tillage. We can do a similar thing with cover crops. Not quite as strong as a relationship there but maybe, actually, we can see that POXC is sort of buffered or plateaued. It takes more tillage to see a reduction in labile carbon, if I am using carbon crops, but it's not nearly as dramatic. If you think about the amount of carbon that comes with manure, in general, versus a cover crop, we can expect there might be some difference there. How am I doing on time Ricardo? - [Ricardo] So you still have like five, 10 minutes. You're good. - Okay, good. So I wanna talk a little bit about the social science aspect of this work. So we wanted to not only look at the data on the field and be able to help growers maybe target their tillage a little bit better, but we wanted to understand how they're making decisions about tillage, so that we can help them make better decisions, essentially. So we used a few different methods. A survey, some social network mapping and some models of tillage behavior to do that. One thing is that we ask growers about their motivations and deterrents for reducing tillage in soybeans and the thing you wanna look at here is on the right column. We've got some state-wide means and basically, if you look at this bold number, for motivations versus deterrents, all in all, growers felt like the motivations for reducing soil disturbance outweighed the deterrents. And if you look at how those flowed, they were motivated, first, by reducing labor. Second, by reducing erosion. Third, by increasing soil health and fourth, by reducing their cost of production. We also looked at how their responses to these different items lined up with their tillage intensity. So are these actually correlated with what they're doing. They say these are motivations but what does that mean for their behavior. And basically, the strongest relationship between the motivation that they reported and their actual tillage behavior was reduced in cost. We also did something called social network mapping. So these were groups, we had these three groups of growers in the different regions of the state and we wanted to understand the relationship between the growers in these groups. So we actually mapped their network so each circle in this map represents one of our growers. The arrows between them, excuse me, represent relationships where the growers indicated knowing one another. The thickness of those arrows is the strength of the relationship. How well they said they know each other. And you can see the different tillage systems based on the color there on the bottom. We also can look at and analyze the network as a whole. So something like degree, the network degree is the number of relationships that a certain grower has. So the number of arrows or vertices that touch one of these hubs. Sorry the vertex, the sphere that represents the grower. So how many people does that grower know within the network? Since we're running a little bit low on time I won't go much more into these, but I just want to show you, this was our northeast region versus our central and southwest. You can see differences in how well connected these networks are. The northeast was pretty tight-knit and most people know each other, versus there are some outliers where people that aren't as well known or don't know other folks as well in these other networks. And then we looked at the relationship between some of these different factors and tillage intensity or how much tillage these growers are doing, and what we saw, again, a lot of numbers, but the basic pattern was that income and the perception that no-till reduces cost, influences tillage behavior in this sample. So as income increased, so the more money a grower has in their pocket, the more tillage they tend to do. So that was kind of interesting that if you've got those expendable resources you're more okay spending them on tillage. Kind of an opposite pattern, the perception that no-till reduces cost, tended to reduce soil disturbance. So if I believe that no-till's gonna reduce my cost of production, I tend to till less. We also looked back at soil quality. Here it captured in terms of CEC, cation exchange capacity. Now if you remember, our recommendation was that on a higher organic matter soil or finer textured soil, which would be a higher CEC soil, you should be tilling more not less. And what we saw is actually our growers in this sample are doing the opposite. And so you could say wow, they're doing something maybe that's irrational or not what would be recommended by the data. But when you talk to these growers you see that actually there could be something going on there as far as tillage opportunity. So these soils that are poorly drained, high-organic matter, also, perhaps, are more difficult to till because you don't wanna till them when they're too wet. You're gonna have more problem than you are benefit in that system. So sometimes gotta think about not just the correlations but the opportunity to implement a management practice. This topic of inference space is kind of interesting. So I showed some data, this bar chart, early on in the talk, and we talked about how, at KBS, they saw a benefit in terms of yield from no-till versus conventional tillage. What was interesting, one way to kind of help validate our findings, is that when we look at where KBS falls out in our different environments, or our clusters, KBS was in cluster E, and if you remember the recommendation, that was an environment where no-till was recommended. So it was kind of neat to see that our findings align with what they saw at KBS, that we would expect, we would predict that no-till would perform well in that environment. But you gotta, what we mean by inference spaces. Just 'cause we do a study at KBS and we show that no-till is successful there, it doesn't mean that when we extrapolate that recommendation or that finding across the landscape, that the results are gonna be similar everywhere. And that goes right back to my first slide about defining tillage and the outcomes of tillage vary, based on the interaction of the tool, the operation and the environmental conditions or the soil in particular, where that operation is carried out. So to conclude, we identified the six unique environments where soybean yields respond more predictably to tillage. Northern latitudes, fine texture, high organic matter soils are associated with the yield penalty for reduced tillage or no-till of eight to 21 bushels per acre. In the northern clusters, yield responds to tillage seem to be about cycling that labile carbon which could be, as we discussed, about warming and drying the soil with tillage. And other research helped support our model as I just showed that situation with KBS. The negative effects of tillage on soil health and other ecosystem services can, perhaps, be buffered by limiting tillage frequency or intensity. Not tilling anymore, or more frequently than you have to. Baseline soil quality. So if I'm tilling a soil that has finer texture or higher organic matter, the risk of losing organic matter, degrading soil quality, is less. And organic matter additions, particularly manure. Grower tillage intensity decisions are influenced by a lot of environmental and socioeconomic factors. So social prominence, network degree. I didn't touch on that but I should mention we saw that growers that were more well connected in their network were tilling more intensively and in discussion with the growers and looking at the literature, what we've come up with is that social prominence is perhaps associated with social risk and that means that if I'm well known in my community and I decide I'm gonna do something innovative, like I'm gonna go out and adopt no-till. If I fail, the risk of that failure's greater for me because I'm well known. If I am lesser known, if I'm an outsider in my community and I go and do some no-till and it fails, the social risk of that is less and so we saw, actually, that growers that are more outsiders in their community, less connected, less visible, tend to be more innovative in the sense that they're more likely to reduce soil disturbance or adopt no-till. A few more acknowledgements. All the growers that we worked with on this project. Again, the team on this project. I wanna thank you all for listening to the presentation tonight and I'm happy to take more questions. Thank you very much.