 Our next speaker is Cami Dixon. Cami is a biologist with the U.S. Fish and Wildlife Service. She spends her time addressing uncertainties associated with managing grasslands and wetlands. She is based at the Chase Lake National Wildlife Refuge in North Dakota, but works across the Prairie Pottle region. She has a program coordinator for the Native Prairie Adaptive Management Program, which was a multi-state initiative focused on managing prairies, invaded with smooth brome and Kentucky bluegrass. In recent years, Cami has also taken on roles to help support pollinator conservation in grasslands. The best part of her job is collaborating with partners and colleagues to address the numerous challenges associated with conserving grasslands and wetlands. Cami's title today is smooth brome and Kentucky bluegrass. Does increased invasion, decreased management, effectiveness in mixed grass prairies? Let's welcome Cami. Thank you. So I first want to acknowledge my co-authors on this presentation. Jill Gannon, who did all the quantitative work that I'm going to be talking about. And then Jennifer Zorn, who does all the data management for the data setup that I'm going to be talking about. So back in the early 2000s, several of us US Fish and Wildlife Service biology, biologists and managers in the Dakotas started to notice that our prairies were looking more like this instead of this. This of course is more of a monotypic type of smooth brome and Kentucky bluegrass. And the other one of course is a high diversity, mixed grass prairie. So as we started having these conversations, we realized that we actually wanted to quantify what was going on on our prairies to determine if what we thought we were seeing was really happening in that we had more smooth brome in Kentucky bluegrass than we previously it had. So we went out and surveyed almost all of our native prairies on US Fish and Wildlife Service lands in the two Dakotas and into Northeastern Montana. So these are the results. This was in the ballpark of 200,000 acres that we surveyed. The y-axis here is percent frequency and the x-axis are our plant groupings. So as you can see our native grass and forb groupings, we have about 25%, 25% or less of the native plant cover. Got a few low shrubs like Western Snowberry, about 10%. But what's most eye-opening on this figure is that our prairies are so highly invaded that we smooth brome and Kentucky bluegrass at the two of them together are making up more than 50% of the composition of our prairies. Got a little bit of crested wheatgrass and some weedy forbs, things like sweet clover as well as leafy spurge. So this was pretty eye-opening to us to see that our prairies were so degraded and we really felt the need to do something immediately to see if we could at least slow down the tide of invasion. So that led us to create what we now call our native prairie adaptive management programmer, N-PAM is the acronym you'll hear me use. At the core N-PAM is a decision framework that uses science and our values to be able to recommend optimal management actions on native prairies that are invaded with smooth brome and Kentucky bluegrass. The objective of the N-PAM program is to increase the native plant cover at the least cost. Here's the spatial area of our N-PAM program. The green polygon, the dark green polygon is the prairie pothole region, whereas the shaded green area are our administrative areas where N-PAM is currently occurring. These purple dots represent our management units and on average we have about 120 management units enrolled in N-PAM every year and we have both mixed grass prairies which is everything to the west of that yellow dash line and tall grass prairies, which is everything to the east of that. So for the sake of the presentation today, I am only going to be talking about mixed grass prairies. So keep in mind we're all, everything I'm talking about is west of that yellow dash line. Sarah Vossick will be giving you a rundown on what's going on in tall grass prairies tomorrow. I call this slide the year in the life of an N-PAM unit. So this is gonna give you an idea of what happens on these N-PAM units every year. So remember N-PAM is a decision support framework. Every year these units are managed and for tall grass prairies, our management treatments are rest, grays, burn and burn grays. So we manage, then we monitor these units every year that management and monitoring data are uploaded into a centralized database and then those data are processed in a pretty complex decision tool that I'm not gonna go into details on today, but I'm happy to discuss afterwards if you want more details. Then that decision tool provides a management recommendation. Okay, so now that you've had a little bit of background on what N-PAM is, I just wanna step back a little bit and talk about how philosophically U.S. Fish and Wildlife Service, how we're thinking about dealing with uncertainties associated with these prairies that are so invaded with smooth brome and Kentucky bluegrass. I think of it as three legs of the stool if the seat of the stool is our uncertainties. So the first leg of the stool is our N-PAM program which I already talked to you about and that gives us our annual decision guidance on how to best manage our prairies. In addition to that, we can do retrospective analyses of the N-PAM data. We can answer questions if we can, we can hope to get answers if we know the questions we want to ask. And then finally, the third leg of the stool is targeted research where we might ask a partner to help address a specific question. And actually a good example of this is gonna be Sean's presentation today. He's giving a little bit later, that's some targeted research we did in collaboration with him. So for the sake of the presentation today, I am gonna be focusing on retrospective analyses of these data, of this data set. So a few sidebars background details on this data set. So the first thing I want you to recognize is a data point for N-PAM is we monitor, we manage, and remember those four management treatments we use, it's rest, graze, burn and burn graze for our mixed grass prairies. And then we monitor again and get the results. So we call this is a single data point, sometimes we call it, it's a single time step. So keep that in mind as you're seeing the results. The other thing to keep in mind, I've already showed you how highly invaded our prairies are. Turns out when Fish and Wildlife Service was first acquiring the lands we managed in the Prairie Pothole region, we thought it was best to idle these prairies for the purpose of a certain group of wildlife. So we didn't do things like burning and grazing for a lot of years on many of these prairies for decades actually. So I always like to point that out, it might not be comparing apples to apples if you're thinking about a prairie that has a long-term history of grazing and looking at our data. So keep that in mind. And then finally, we use what's called the belt transect method to do our annual monitoring. The belt transect method is a fairly rapid assessment. It's a lumping versus a splitting monitoring protocol. And we do by percent cover and we lump plants by functional group. So keep all of that in mind as we go into our first retrospective analysis that I'm gonna talk to you about. So the first analysis that we did is relative importance and we defined that as the amount that a variable explains the change in smooth brome and Kentucky bluegrass cover relative to other variables. Okay, let's get a few more details on this and talk specifically about what variables we looked at. So in terms of doing this analysis, we could only pick variables that we had data for and we wanted to pick variables that were important to us. So these are our base variables that we hope, these are our base variables, seven base variables and they are treatment. Remember those four mixed grass treatments I've been mentioning. The dominant invader, whether it's smooth brome, Kentucky bluegrass are a combination of the two. The defoliation history for NPAM purposes, we look at what happened the prior seven years in terms of defoliation. The invasion level, this is kind of that threshold thing that's already come up in some previous presentations. The prior year precipitation, the long-term climate which includes both precip and temperature for prior 30 years and then the ecological site. So these are the seven base variables. We also did interactions with these. And so truly there are in the ballpark of 30 variables but I'm not gonna show you all that because that gets pretty overwhelming to look at in a presentation pretty quick. So we're gonna focus on these. As I start revealing the associated number with these, hopefully you can see that at the bottom but let me make sure I read this clearly. So these numbers are gonna range between zero and one. And so what these numbers mean, if it's a zero, it means that that variable doesn't explain the change in cover of smooth brome and Kentucky bluegrass much. Whereas if it's a one or near one, it explains the change in the plant cover a lot. So let's look at these and see what we learned from this. Okay, so the treatment. Now again, remember the question we're asking, how do these variables explain the change in smooth brome in Kentucky bluegrass cover? That's a question you wanna ask in your head as I go through all of these. So treatment, treatment's an important variable for explaining the change in the cover of those species. How about the dominant invader? Well, it's pretty important for smooth brome, not at all important for Kentucky bluegrass. Defoliation history, somewhat important for smooth brome, not at all important for Kentucky bluegrass. The invasion level, it is somewhat important for smooth brome, very important for Kentucky bluegrass. Prior year precipitation, really important for smooth brome, not important for Kentucky bluegrass. Long-term climate, important for both of them. And so is the ecological site. A few take home messages on this slide. This was actually pretty heartening for me to see that treatment is this important in terms of explaining the change in the cover of these two species. That's something we can actually control, theoretically, whether we're gonna go out there and burn or graze or even rest as a treatment too. So in comparison to things like long-term climate, which of course we have no control over or prior year precipitation or even ecological sites. So I was actually pretty excited about that. That made my day when you spend a lot of time working with smooth brome and Kentucky bluegrass. The other thing I wanna point out about this is note the differences in these variables as you go from smooth brome to Kentucky bluegrass. These variables have, some of them have similar importance, but they're different from one another. So essentially I think of these two as they're two different beasts, even though a lot of us are challenged to manage prairies that have a combination of the two. One thing I really wanna emphasize, so this suite of variables that we used for this analysis, again, these were variables that we had data for, certainly variables that we're interested in. But for example, if there's something going on, like Dr. Lam talked a lot about the nutrient cycling and that sort of thing, if there's something like that driving the change in smooth brome cover, we didn't capture it here. I just can't emphasize that enough. This is based on the variables we had available to assess. So I could go through each of these variables and show them the associated figures that go with them, but I decided to select just one of them. This whole concept of this invasion level, this is a bit of this threshold conversation we've already talked about and I chose this one because it seems this is something we talk a lot about, especially in the coffee shop, after hours, campfire chats and so forth. So I'm gonna show you the results of the analysis, the specific graphs that went with this relative importance analysis. So just bookmark that 0.48 for a smooth brome and the 0.99 for Kentucky bluegrass. Okay, here's a few details of the analysis. So the question we asked to the data is do burning and grazing become less effective as invasion increases? Well, what's your hypothesis? Our hypothesis was as invasion level increases, burning and grazing will decline in their effectiveness. Anyone agree with that that you would think that's a reasonable hypothesis? So here is our sample size. We had 105 units, data from 2010 to 2020. This is 843 data points. Now, do you remember when I told you what a data point was earlier? We monitor, we manage and we monitor again. That's a data point, that single time step thing. So keep that in mind. We had 334 rests, 367 grazes, 90 burns and 52 burn grazes. So let's start to look at the graphs. Okay, so we're gonna start with smooth brome and I'll walk slow this first graph so you can get an idea of these box plot diagrams if you're not used to looking at them or it's been a while. So we're talking about smooth brome here. So our question remember is do burning and grazing become less effective as smooth brome increases? And remember our hypothesis is that yes, that's gonna be the case. So on these box plot diagrams, first of all, smooth brome is on the Y axis for percent cover. Across the top here, we have our level of invasion in three categories on three different panels. So medium is our average invasion level for all the units that were part of this analysis. So that's average invasion of smooth brome and Kentucky bluegrass. On our respective units, keep in mind. So anything above average would be high, anything below would be low. We have our management treatments on the, excuse me, we have our management treatments on the X axis here, of course rest, graze burn and burn graze. This dashed line on each panel, that dashed line is the starting state of the smooth brome cover before we applied the treatment. So the results you're going to see is anything above that dashed line is an increase in smooth brome cover. Anything below is a happy face. That's to help you remember that we want below that cover. So I put a happy face just to help you remember that and help me too. What's really confusing is if you suddenly have native plant cover on the Y axis of these, then you have to, it takes me a minute to flip my mind around. So I kind of like the happy face. Okay, so let's start looking at some results. For rest, it's predicted that under low invasion, we're going to see an increase in smooth brome cover when we rest. Same under medium and same under high. So this is not living up to our hypothesis so far in terms of we should have, I would have expected rest under low invasion to potentially be at the maintaining or even decreasing potentially depending on. So all right, let's look at grazing. Okay, under low invasion, it's predicted that a graze is going to increase the smooth brome cover. Under medium invasion, maintain. And under high, roughly maintain. Again, not meeting our hypothesis and there isn't, is similar to rest, it was pretty much across the board among the three panels. In graze, there's a very slight difference from low to medium and low to high, but not dramatic across those three invasion levels. Okay, let's look at burning. Okay, before we look at burning and these box plot diagrams that taller the box, the less confidence we have in the results and the longer the whiskers. So you can see, just be wary of that. And when I pull out burn graze, you'll notice it pretty dramatically. So under low invasion, it's predicted that if we burn, we're gonna see an increase in smooth brome cover, potentially maintain under medium and potentially maintain under high. Again, my confidence on this high with the long whiskers and even on the medium is sort of shaky on those. And then finally, the burn graze. As you can see, these are real tall boxes and long whiskers, but burn graze is predicted based on the data we have is predicted to show an increase in smooth brome under low and medium and possibly maintain under high. Okay, so what does this all mean? I would appreciate feedback on this if you all have this because I'm still getting my head around it, but bottom line, we are not seeing when you look, especially at, and for now I'm just gonna, I'm gonna put the burn graze out of my mind just because I don't have a lot of confidence in these results with the high variability in there. So let's just talk rest, graze and burn. So with rest, not much of a difference across the board. Burn, there's not much of a difference either. If you lined up these red boxes where they're sitting on these dash lines with each other, those red boxes are crossing over there. So there's not much of a difference across those three panels. There's a slightly something going on in grazing, but it's not too dramatic. So that said, all of this, remember what the relative importance of smooth brome was? It was like 0.48, it had some importance, but not as dramatic as Kentucky bluegrass. So again, there's not as much variability as you're going from low, medium and high, especially for rest, graze and burn. Okay, are we ready for Kentucky bluegrass? Okay, same question. Do burning and grazing become less effective as Kentucky bluegrass increases? Okay, so now remember our Y-axis is percent Kentucky bluegrass cover. So first of all, under a low invasion level, arrest is predicted to increase Kentucky bluegrass. Under medium increase as well, and under high decrease. Boy, I'm dying to hear some of your thoughts on that because this is really a bit mind-blowing on this, but bottom line, this is pretty much opposite of our hypothesis in that rests, rests do better under high invasion levels. Okay, let's see what grazing did. Quite similar to rest as you can see. So grazing is predicted to increase Kentucky bluegrass under low, also increase under high, but look at the, under medium, I'm sorry, but under high it's predicted to decrease the Kentucky bluegrass cover. Let's check out burning. Okay, so burning under low is predicted to increase the Kentucky bluegrass cover. Under medium and high, both predicted to decrease Kentucky bluegrass cover. And then under burn graze, the burn graze is really trying to follow our hypothesis. Again, with these long bars, a lot of variability in there. We can't quite say that yet, but that's really trying to say, yep, a burn graze becomes less effective. Looks like we show a decrease in Kentucky bluegrass under low, potential decrease under medium, but under high, we're gonna see an increase. So all of this, there's a lot more variability, especially compared to the smooth brome as you go from low, medium and high. So what this means is that this is, remember this is a 0.99 for the relative importance, is that this is an important variable for Kentucky bluegrass because there is variation across these three invasion levels. Okay, let's talk about a few take home messages. First of all, management treatments are important. As I said, on that relative importance table, I was pretty excited when I saw that treatment was a one. Oftentimes our managers and especially our administrators will say, well, Kami, why are we spending money on prescribed burn? Alls you're doing is saying we need more fire on the landscape, you want more money to burn, you want, and they say, how do you know, how do you know this isn't all driven by climate, that long-term climate prior to your precipitation or ESD isn't driving all this and our management treatments aren't doing anything? Now I have this analysis to show them that our management treatments are important. So that's really important to me. And I think it's important to tell our managers, our land managers, that keep doing what you're doing, keep burning and grazing. So also, if you remember in the relative importance analysis and as well as the invasion level, figures I showed you, smooth brome and Kentucky bluegrass need to be, they're their own beasts, I like to say, in terms of you really, of course our goals to increase the native plant cover, but you really need to know what your dominant invader is, the species that you want to manage against. This is this threshold question. Management effectiveness did not decline with increased invasion as we had thought. Really, there weren't many, I think maybe there was one scenario under burn graze where it was really trying to follow our hypothesis where to show that under high invasion, our treatments aren't doing a thing. That didn't really occur. And again, in this world of working with smooth brome and Kentucky bluegrass, that's actually a little bit exciting to know. And then another take home message, keep studying and learning. We have done this one analysis and I talked you through how we did that. So we're gonna look at this a little bit different way, this threshold or invasion level question. For the NPAM program, we have four native plant cover states. The first is zero to 30% native cover. So the question we're gonna ask next for our next analysis is if you are a manager and you have a prairie with zero to 30% native plant cover, and that's in 2010, in 2020, have you moved to this second higher state to that 30 to 45% native plant cover? So with that, I want to emphasize that this is, I appreciate being here and having this opportunity to be able to talk through these and think through these. And these are challenges that don't seem to go away or don't seem to get any simpler as I was watching or listening to Dr. Lam this morning, I'm like, oh my goodness and trying to get your head around all that and bring all this together. But I do have some optimism with some of these results. Definitely we need to keep going, we need to keep burning, we need to keep grazing. So with that, I'll answer any questions you might have or try to answer anyway.