 So we'll start out with our first speaker today, and I sure it's a pleasure to have Dr. Eric Lam here. Dr. Lam is a professional in the Department of Plant Sciences at the University of Saskatchewan. He holds his bachelor's of science degree from the University of British Columbia and his master's of science from Lakehead University. He received his PhD in grassland ecology from the University of Alberta. He has a plant ecologist with expertise and research interests in fields, including rangeland ecology and management, fire ecology, statistical ecology, but I don't even know if that is. It's a plant and soil interaction. So Dr. Lam's talk today is Roots Matter. How Smooth Broome alters the structure of soil microbial communities and soil ecosystem services. Please welcome Dr. Lam. Great, thank you. Well, thank you, Kevin, for that introduction. One of the things you're gonna see in this talk here is a lot of statistical ecology. So be warned. So I really appreciate the invitation to come here. This is actually my first time, even though Saskatchewan's pretty close to North Dakota. It's actually my first time spending any time in North Dakota. And I was kind of sad as I was driving down here that there was a lot of snow around because it looked like there would have been a lot of nice places to stop and hike around along the way. So I'm gonna be talking here about work that has been done in my research lab over the last 12 to 14 years. We've been quite interested in us in Smooth Broome as it's a problem in the US. It's also a problem in Western Canada. And we've been really interested in understanding some of the mechanisms underlying how those invasions happen and why species like Smooth Broome seem to be able to hold on so tightly once they get into a place here. So obviously, something like this, there's gonna be a lot of other people who've contributed to this work. So I'll acknowledge them right off the bat. One, Steve Siciliano. He's a professor in soil science. He's a soil microbiologist. So he was the leader on some of the microbiology stuff we talked about here. Dr. Steve Mamet. He's our statistical ecology wizard who helped us do a lot of the stuff we've done here. Jen Bell, a PhD student recently finished who's done the most recent work in our group on this topic. And then Candace Piper, a master's student who produced as part of her master's the most productive data set I have ever had and the pleasure of working with. I think Candace is up to 12 publications now off of that data set. The two from her master's and then we've just kept reusing it because there was so much good stuff in that data set. So something for those of you who are students to aim for. Okay, so you probably don't need any introduction to smooth brome at this meeting here. Of course, it's a forage grass. If you're raising an origin, in Saskatchewan we had a widespread invasion going back to at least 1920. There's lots of records of road sides being seeded with smooth brome in the 1920s and into the 1930s. It was the go-to species for the Ministry of Highways for a very long time. And so we basically have a nexus of invasion every mile on the grid roads throughout the province. Of course, it invades into intact grassland which is why we care about it from a conservation perspective. Obviously ranchers still it's a very useful species when your primary goal is to be feeding cows but from a conservation perspective it's a real problem. And there's some work from Alberta that shows that as that smooth brome invasion progresses we're looking at a 70 plus percent decrease in native plant diversity where that invasion rolls over. And my next project in this area is actually to see whether that actually qualifies some of our prairie species to be species at risk because that level of population decline from an invasion should kick in the IUCN criteria for a lot of different plant species. So it's a problem and we've been seeking to understand how brome functions and why it's so tenacious once it gets in. So we've been working on this from this perspective for a long time with this particular field site undergoing a small ungrazed piece of grassland being our primary study site. So for context, this particular study we're about 120 kilometers south of Saskatoon. So Saskatoons where I'm based at the University of Saskatchewan. So here we're right on the margin of mixed grass prairie and fescue prairie. So fescue prairie being of course the primary grassland in the Aspen Parkland which is what we have all around at the northern edge of the Great Plains here. So lots of agriculture here. We're looking at anywhere from 70 to 90% land conversion for agriculture. So we got pretty small remnants left. And this particular spot happened to be a remnant because it was where two roads came together in an angle and there was a little area that nobody felt the need to plow up I guess. So little remnants and, you know, it's a, oops, there we go. And as a for scale then it's a, it'd be about a day to get down to us here. A little bit longer because the roads weren't in great shape. So this particular site, what Candace did, she went out and she found places where we still had intact native and we had places within the grassland where the invasion was happening. So we had everything from completely native to places that had an ongoing invasion. So 20 to 50% brome, 50 to 80% brome and then spots where it had gone to 100%. It's likely what was happening here, it was probably golfers were the problem here that the gopher mound disturbance was where the brome seemed to be seeding in on. And then we would get expanding patches going out from those little ground disturbances as there was obviously a big seed source just along the roadside nearby. So that was the studies design. So very, very simple design and we started measuring the heck out of it. So started with traditional plant community stuff and we see what is kind of the expected results of a brome invasion. So these figures here are just all brome percent cover in the plots along the X axis and then this is total species richness on the Y axis, species evenness on the other one. So exactly what I'm sure you're familiar with when all of your sites as brome cover increases, we see fewer and fewer species. As brome cover increases, we see evenness going way down. So we're the native species that are present are usually one or two individuals in among the big mass of brome. So exactly what we see in most places, this is why of course brome is such a conservation problem. Then we started digging into the soil and we started looking at all of the ecosystem components here. So one of the first things we asked was, well, so that's the above ground impact of brome. It's native prairie, 80, 90% of the biomass is below ground. So we started looking at using some molecular techniques to identify plant roots. We started looking at how the relationships between above and below ground diversity. What we see here, probably no surprise is that as above grounds in diversity increases, we also find a greater diversity of root species below ground. So if we flip this plot around in our mind, of course, as brome becomes more abundant and our species diversity above ground is going down, our species diversity below ground in the root community is doing the same thing. Interestingly, this pattern extends all the way into the B horizons. So we're still, it's obviously a bit weaker, but we're still seeing that brome impact on the below ground diversity as we go down below ground. Other things that we see, of course, we see increased productivity. This is why brome is such a good forage grass. It's really productive. So as brome cover increases, we're seeing greater plant-standing biomass and we're seeing greater litter accumulations. So again, I don't think I've shown you anything thus far that you would find surprising. Anybody surprised so far? Okay, that's good. I'm just trying to set the stage here thus far. So then we started asking questions about how the soil is functioning and who's acting within the soil. So here what we looked at was in fully invaded locations versus native locations. So this is the most extreme comparison at the site. We looked at nitrogen mineralization and nitrification. We see that under brome, we've got much, much higher nitrogen mineralization happening. It's not extending over to nitrification. So what that's telling us, we're getting faster nutrient turnover happening under the brome and those nutrients are being taken up quite rapidly, mostly probably by the brome. So there's very little of that mineralized nitrogen going into the nitrification pathway. So it's being sucked up by the system and what we would suggest there is brome is probably enhancing its hold by speeding up the soil nitrogen cycling and then grabbing that nitrogen as soon as it's becoming available, pushing it back into new above ground biomass. So we're speeding up what's happening in the soil and brome's using that to its advantage to sustain that high plant productivity. This is probably also due to the fact that we've got the higher amounts of litter happening under brome. So we're more substrate for mineralization to be working on. So then we started diving into the microbiology. So we went, we took soil samples. We used a next generation sequencing to identify not only the plant roots, but also the fungi, the bacteria and the archaea within those samples. In both the A and B horizons and in the soil. So this is a summary figure here showing our initial analysis of just the bacterial data. So very simple metric here. We're just asking soil bacterial diversity. And then we're using a structural equation modeling approach to ask what are the mechanisms within this community that are driving soil bacterial diversity? Because you will see lots of suggestions in the literature that if the plant community goes to a monoculture, that's probably going to simplify or reduce your diversity in the soil community. So we put in all sorts of variables that we thought might be important. Obviously you got above ground things, plant community. You got to consider litter, amount of root biomass in there, things like soil carbon, soil nitrogen. And I see which of these patterns is affecting soil diversity. So at the back end of the model, for those of you who aren't familiar with structural equation models, these are testing causal networks where we're, for example, saying increased smooth brome increases the amount of litter, which decreases the overall plant community richness, which decreases root C to N ratio. So we can kind of track causal pathways through the system. And we expected when we fit this that we would see something pretty simple where probably litter C N ratio would be the driver of bacterial diversity. We couldn't get the model to fit until we put in an ad hoc relationship to get it to fit, which is this pathway right here. And so if you've used structural equation modeling, you'll know that if your model isn't fitting and you have to add in something ad hoc, that's telling you that you've basically got a theory gap. We had no reason to think that there was a mechanism linking brome and soil bacteria that didn't involve these variables. And because we had to put this pathway in, it's telling us that brome is doing something independent of all of these other variables. We spent a long time trying to figure out what was going on here. Because the other thing that was happening was brome was causing an increase in bacterial diversity. So not the pattern we're supposed to be expecting. It was going to a brome monoculture was increasing the diversity of soil bacteria. That doesn't make a lot of sense, at least based on the literature at that point. Finally, what it realized was happening is brome was selectively suppressing a small number of very dominant bacterial species within that soil community. So in this figure here, the black bars are showing basically relative abundance of different bacterial species in the native community, and then the red is invaded. What we see, some of these really common ones in the native community are being selectively suppressed by brome, and that's actually producing what's probably a competitive release of a whole bunch of rare bacteria in the soil. We've since in other work, going to things like canola and stuff like that, where we've got a much simpler system, we've demonstrated in a number of different plant species that it's quite common that plant species will be selecting for rhizosphere bacteria kind of at the species level. So this isn't an unusual mechanism, but it is kind of surprising, and probably why the received wisdom before this was diversity goes down, when brome goes down, is the previous studies were done using less sensitive techniques. So we're using next generation sequencing, which gets much deeper into the community, better able to pick up the rare things. So we kind of left it at this point for a while, Candice graduated, went on to a job and started a family and things, but we were still thinking about what was going on in the system, why this was happening. And we had collected at the same time as we collected these, the bacterial data, because once you extract samples, it's easy to amplify another region, test a different, go after a different group of organisms. So we had, when we'd collected the samples, we'd also run the archaea and the fungi. So we kind of sat on this data set for a while because to be honest, Steve Ceciliano and I, we didn't really know what to do with it. Cause at this point, we're dealing with each of these bacterial data sets. We've got about 120 samples. Each of these data sets has anywhere from 200 to 500, 600 fairly common bacterial species in them. So kind of separate matrices. Plus we all have all that environmental data. We weren't sure statistically how we could do anything sensible with it. Steve Mamet came along as a postdoc who is a, just a wiz at this stuff. And we started thinking about how we were going to, how we could actually tackle this data. And if you love your complicated statistics, this paper will keep you happy for quite a while. Basically the approach here is to, so we've got all of these different community matrices. We have to simplify them a bit. So there's a lot of noise in sequencing data. So we have to get rid of some of the, some of the noise, try and enhance the signal a bit. We've got our covariates. So basically the brome gradient, our plant community stuff. And then we need to start working through it to identify who the important players are within each of those data sets and how they're interacting with each other. So trying to identify basically the environmental covariates that seem to be associating with heases of the microbial community. And then we need to identify the key interactions within the microbial community that are important. So when you've got, let's say 500 different different bacterial species, you know, you can do the math and how many potential pairwise interactions there are between species in there. So the challenge is to dig through that mess to identify pairwise interactions that seem to be important. So that signal we saw earlier where suppressing three or four common bacteria competitively released a whole bunch of others. We're kind of looking for those kinds of signals where a change in abundance of one is associated with a fairly radical change in abundance of another species. And so with that, by kind of simplifying the data set down where we get to what we think are the key players within the microbial community and we get to what we think are the key environmental variables. Then we can go back to our kind of structural equation modeling approach asking how these are relating and how they might be driving ecosystem function. So this is a very complicated diagram. I originally had three more associated with it but I felt we could probably keep it at just this one. This diagram here is just to show you some of the data and some of the process. So some of the key variables we've identified in here are below ground, soil organic carbon, probably no surprise that that's gonna be a major driver of microbial community dynamics. Root biomass turns out to be a pretty important driver. And as we begin to break it down, we see these arrows here indicate some of the really important brome microbial dynamics that are going on. So the amount of brome root biomass that's present, it drives things like soil organic carbon. Turns out soil organic carbon is fairly important for both the fungi and bacteria, root biomass. Same thing, root biomass and soil organic carbon, of course, are gonna be fairly tightly correlated with each other. So kind of from the brome to bacteria and brome to fungi pathways, they seem to be driven by the amount of brome roots that are present. On the other side here, we found that the nitrogen in the above ground litter is pretty important as well in driving both bacteria, fungi and the archaea within the soils. And if you dive in further into these figures, each of these points is a sample, they're color coded by the amount of brome that's present. So the redder the color, the more brome there is present. And so you can see there's all sorts of significant correlations happening within these. But the important point is that we've got, we're kind of down to these two major drivers. And then within the soil community, turns out that archaea are actually a pretty important player. They're not all that abundant. There's not all that many species of them relative to bacterial and fungi, but the archaeal community appears to be structuring both bacteria and fungi. Bacteria are also important in structuring the soil fungi. So we've got these very complex dynamics among very, very different sets of organisms, ultimately appearing to drive the fungal community. So maybe we made a bit of a mistake by focusing on bacteria in that first paper. Maybe we should have done the fungi first. So I hope you can kind of see it out, like this is a complex web of interactions happening here and involving well over four to 5,000 different species. So few of the key actors that are happening here driving this, we can get down to naming a relatively small number of microbial species that seem to be the drivers of all these relationships. So most of the species that are present are just kind of passengers along for the ride. There's a small number that are driving it. We actually see that it's a relatively small number of plant species that are driving these relationships as well. Obviously, brome is important, but once we looked at the root diversity data, what it turns out is particularly for the fungi, it's not the presence of brome so much that's the problem. It's the fact that when bromes present, a lot of the native species are disappearing is the problem, particularly for the fungi. So it's that displacement of native species that's driving some of this, not just the presence of a new really productive species. So then as we get to trying to put this all together, we see that we've got these bacterial mediated relationships structuring fungi driven by brome shoots and brome litter in the soil A horizon. And then when we drop to the B horizon, it actually becomes the brome roots that are a little bit more important. So every analysis we were doing here, we had pairs of samples, one taken from the A horizon, that top 10 to 15 centimeters in the charnism and then samples down to about 30 centimeters in the B horizon. What we're looking at here now is a kind of a summary of all this. So we've got brome shoot biomass, brome root biomass, archaeal diversity, bacterial diversity, fungal diversity. When we're looking at the coefficients on these lines here, so the coefficients that are ahead of the little slash there, those are what's going on in the A horizon, the ones after the slash or the B horizon model. So we see in the A horizon, we see brome shoots and brome litter, really important driver of bacterial diversity through that selective suppression. Down in the B horizon is exact opposite. So brome roots don't really matter in the A horizon, significant driver in the B horizon. The archaea seem to, for the most part, once we come down to it, they're not really responding to the brome, at least in this particular model. So we had to run this model multiple times using different sets of important players. So all I'm showing you here is the bacterial, or sorry, the diversity model. And then we see this bacterial diversity, driving fungal diversity, particularly in the A horizon and the B horizon doesn't matter quite so much. So at this point, we're feeling like we have a little bit of a handle on this. So you should be aware that kind of underlying this, we actually had to write a stats paper to demonstrate that some of these techniques actually worked. So say, fortunately, we were able to have Dr. Mamet on the team for this. But at this point, this is our best take on what's going on. So this point, we started getting uncomfortable with the fact that we've been mining the same dataset from the same site for quite a while now. So what if we had a really weird site? It's always niggling in the back of your mind that at some point you gotta demonstrate that this is going on somewhere else. Now, one of the challenges with getting all of this stuff published is, of course, we run into a lot of reviewers who ask for multi-site, multi-year data. And we would lead that, took a good portion of our lives just to get this data from one site. So, be kind to us. But at some point, we do need to kind of expand out. So we decided to go with a different site, one right on the edge of Saskatoon, where we could have ready access to the site. It's a piece of native prairie owned by the University of Saskatchewan, 15 minutes from the main building on campus. So we're like, okay, we're gonna do this right. We're gonna sample across a wider spatial area. We're gonna sample temporally. So we're gonna sample from green up to freeze up, everything else, because this site's super close. Actually, this is actually a very old photo of this particular site. I actually, this is a neighborhood now, which causes all sorts of trouble when I start doing prescribed fire out here. But I keep this photo around to show that we were here first. So, and so this particular site is against a fescue grassland. It's actually one of the largest pieces of intact fescue grassland in Saskatchewan. And we started, and we also have multiple invasions there. So at this particular site, Brom is a problem. Kentucky bluegrass is actually a bigger problem. Plus we have absent, and we have perperanial south thistle, and we have, what else, we have Canada thistle, knotting thistle, you name it, we've got it. So we've got in a multiple invasions happening at the same time, plus our Brom. So we went out there and we sampled weekly from say from green up to freeze up. We had a very, very committed group of undergrads and technicians who were out there every week, regardless of the weather. And we relented on the last day when it was a little bit of snow coming down and the ground was starting to freeze. We're like, okay, you don't have to go out next week. And we sampled about 500 locations over that year, measured plant community, all sorts of ecosystem services, soil community. So again, archaea, fungi, bacteria. And we actually collected this data set in 2014. And we were a little overambitious, shall we say. So the sampling design, again, very complicated. And it was only once we got a really talented PhD student, Jen Bell, in our research groups that we were actually able to tackle it. And she finished her PhD in 2021. So this is fairly, fairly recent stuff. So important thing here, this is just ordinations of the plant community. And I've scored the sample locations based on the abundance of the foremost common invasive species. So you can kind of see we have, we have Brom invasions, bluegrass invasions, thistle invasions, south thistle invasions. They overlap to some degree, but not completely. And then we have lots of locations within the site as well that are completely native. So we started, and then we measured soil ecosystem services. So we decided to take an explicitly ecosystem service approach here, where we measured a bunch of different processes, mostly chosen because there were things that we could readily measure. So for example, Steve had a detector in his lab that led us for just a few cents a sample, measure glyphosate degradation on a little soil sample. So that gives us a measure of ability of the soil to handle potential contaminants. And this would be just the data from that one ecosystem service over time. So this isn't quite Julian day. This is April to November. So we have this for everything from forage production, things we can classify as soil conservation measures water purification measures, soil climate regulations. So these would be things like soil organic carbon and then measurements of nutrient cycling rates. So basically everything we'd measured in those previous studies plus a bunch more stuff. Some of the really surprising things we found is number one, the invasion actually was having really limited effects on ecosystem services. Probably not what we wanted to hear, but that's what the data showed. And so what we're looking at in all these figures, these are at kind of peak productivity in the season. So this is basically late June to the beginning of August. There's another figure for spring, another figure for the senescence period, but this one mostly makes the point. So when we look at things like forage production, we've got the green ones are highly invaded. So these are where more than 50% of the plant biomass is an invasive species of some kind. The middle ones there, those are ones where invasive species are present, but they're not making up the majority of the biomass. And then 100% native, no invasive species present. So first thing we see, well forage production is up when we have an invasion. Okay, no surprise, we've got Kentucky bluegrass and brome are the two main invaders. So no surprise, we're gonna see any changes in that. Water purification related ones. So this would include the glyphosate degradation, for example, no effects. Same thing, climate regulation, we're not getting any differences in soil organic carbon or anything like that with invasion. Some of our soil conservation measures, we're seeing a little bit of an impact here. I'm not quite sure what's going on with these ones. And then nutrient cycling seems to be highest when we've got that kind of mixture of invaded and native community present. So a little bit surprising, we were expecting to see quite a bit stronger effects to be honest on this. And we do, I'm not gonna show you the data, but we do see all the standard effects of the invasion on plant community diversity with one very interesting exception. Native plant diversity is actually higher when brome was present. Again, kind of scratching my head on that one. What it turns out actually is this particular site, we do conservation grazing where we graze at an extremely late stocking rate for about two months of the year. What it turns out actually, well, the cows are doing a really good job of keeping the brome under control. And so most places where brome is present, they're keeping the brome clipped down to just a few shoots and they're preventing it from going to seed. And they're kind of digging around with their hooves a little bit. So that's where we find all of the low-growing native forbs. We find lower diversity here in rough fescue, which it goes to a climax where the community is 90 to 100% rough fescue. So there may be some weird things going on at this site as well. But as we try and put the data here together, we start seeing a feedback loop where as in the previous study, when we get invasion, and remember here now we're invasion writ large, so this is not just brome, but bluegrass and other species, we start seeing altered assembly processes within the microbial community. So one of the really neat things that Jen did in her thesis was she started trying to classify the mechanisms that were structuring the soil communities. So is it internal dynamics among the bacteria that's structuring bacterial diversity or is it the ability of new species to arrive at the spot? So kind of external versus internal mechanism structuring the community. She found, yes, invasion is restructuring these assembly processes, particularly for fungi. This then is altering our soil microbial communities and we are seeing some limited feedbacks into ecosystem services and feedbacks back to the invasion themselves. So we're seeing some of the same processes I showed you from the previous study between the altered microbial community and the invasion. But overall the impacts are surprisingly limited given what we might expect when you take a new novel plant species and they're going to 60, 70, 80% of the total biomass in the community. The impacts are actually surprisingly small once we get to this kind of course ecosystem services level. So some of the key messages I'd like to leave you with here is number one, we see right from the beginning smooth brome is altering those soil processes in ways that are maintaining the invasion. So I think it's that input of extra litter as the brome begins to grow and then the higher mineralization rates that are being driven by that extra litter is probably one of the critical mechanisms keeping brome present. Because remember that nitrogen that was being mineralized was not going into the nitrification pathway. So it was being sucked up by the plants that were present very, very quickly and probably turned back into brome biomass. So there's kind of that positive feedback loop going on with brome right from the beginning there. And then this cascades into a fascinating series of dynamics within the soil. But if we're gonna take kind of a practical approach to this, I mean, there's not really any way that we can be manipulating those soil processes. We can only really manipulate the soil processes by what we can do to manipulate the plant community itself. The microbes are going to do their own thing in response to what the plants are doing. And so I would suggest that probably our key with the brome is we wanna disrupt this cycle. So preventing that accumulation of excess litter, preventing that on a jump starting of the nitrogen cycling, that's going to, if we can prevent that, that's going to put those slower growing, more conservative native plants on a more even footing against the brome. And I think that's actually what we see in our current and prairie example where in the end, in this case, using a light conservation grazing to prevent that litter buildup combined with a regular prescribed fire regime, we're actually, we're not getting rid of the brome. It's probably there to stay, but we're keeping it down to the point where it's really not causing us any serious trouble. And so I think if we use that lens of recognizing litter as that key point in the cycle, what can we, and then ask yourself, what can we do to at that point? I think that that's where we can actually have impact on kind of controlling brome invasions. So that is my talk. Thank you very much. Thank you for the invitation to come here.