 So welcome everybody. It's a great pleasure to introduce Daniel Amor, who is a postdoc at MIT. Daniel is one of these brave physicists that has been trained in physics and then moved to do also experiments in biology and try to combine theory with experimental data. Then it started in Barcelona physics, and then these PhD in experimental sciences and sustainability in Giroa, and then after a postdoc with the consulate in Barcelona moved to MIT four years ago. And today is going to talk about how microbial communities responds to perturbations. So before I leave the floor to Daniel, I just want to remind the rules for these webinars. So there are no strong rules. You can typically, you can just unmute yourself and ask a question during the talk, if that's fine with Daniel. So just ask a question during the talk, we try to keep it as interactive as possible. So with that, thank you very much Daniel for accepting our presentation. Thank you Jacopo. It's a pleasure for me to give this seminar today and I'm looking forward to the discussions after a while. Before I begin, am I sharing the screen correctly? Is this the, are you now seeing my, okay, because yeah, it's, it doesn't appear as when we tried before, but if it's working, it's great. Okay. How do microbial communities respond to perturbations? I'm here today to speak about this question and I wanted to first begin by looking back at the first time in history in which this question made sense at all. And our story begins in France back in the 1860s. At the time they was, there was a huge crisis in the wine industry. And many, many bottles of wines were spoiled after storage. And it was so bad that Napoleon itself called repastor and asked him to find a solution for this problem. At the time repastors has already shown the world that fermentation was not the purely chemical process that everybody had in mind. Instead, tiny cells or tiny yeast cells were responsible for the fermentation of the precious alcohol. So this time he's facing the challenge of saving the wine industry. And for the first time in history, he takes a lot of samples of wine and starts looking at wine at the wine and under the microscope. And here's what he finds out. So it was usually easy for the winemakers to obtain the alcohol because from the group juice, the yeast that already were leaving in the skin of the grapes start their activity and they produce, they turn the sugars into alcohol. And after the fermentation, they naturally decrease their activity and when everything goes well, the wine can be stored and it will taste good for a long while. However, he also found out that in some cases one of the microbes that was already in the wine would take over after fermentation and will turn the wine into a sour wine. Or in other cases, it was another microbe that will turn it into a cloudy wine or a slimy wine. And for each of the different maladies of the wine, he was able to find a particular microbe that was causing that malady. That was already a great discovery at the time, but he obviously didn't stop there. He was looking for a way to prevent this. And he, after a lot of experiments, he found out that this was the best thing to do. It was a perturbation which consists in increasing the temperature of the wine just after the fermentation up to 50 degrees. When you do that, the abundance of the microbes in the wine goes down very significantly and it was enough to, in most of the cases, prevent the emergence of any of the maladies. So this was actually the origin of fermentation. Many of us might know, at least I thought before that fermentation started with milk, but actually it didn't. It started with wine and then it was transferred to other processes. But if we look at it from an ecological perspective, is one of the first cases in which somebody intentionally drive and engineer the state of a microbial community. So from the different stable states that the system could reach, Pasteur was directing it into the most desirable state, which is the good wine. And today we know already many different ways, the good ways of producing wine. However, when it comes to predicting the response of a complex microbial community to a perturbation, we are still pretty unsophisticated. So let me show you an example related to a complex community that is the microbial community living in our gut. So here we have a study in which the scientists were tracking the composition of the microbiome of a human individual. And the different lines shown groups of important taxa in this community. So what we see is that four weeks at the beginning, the composition of these groups is pretty stable. But one day this individual got an infection. I think it was Salmonella. And from this infection, you can see how one of the groups goes up because it's the pathogen belongs to this group. But then in a process of about two weeks, they was clear. However, in the during this process, there was an alteration of the abandon says of the other microbes that responded to this change in the community. And the study continues for many, many weeks after and we see that it reaches a different kind of stability. So the composition doesn't look the same as it was before the infection, but it suggests that it has reached a different stable state of the community. However, in these cases, the community is so complex, and we have so little control and limited ability to do experiments with the human gut community, obviously. So overall, it's very difficult to disentangle the mechanisms that are driving these transitions. And beyond human health, we have a huge problem nowadays with the human activities and climate change challenging our ecosystems. And in many cases, we know that ecosystems can also display alternative stable states. Some of these stable states are also linked to microbial activity and one good example is the one of lakes. So here we have a picture of a lake in its oligotrophic state in which the Newton concentration is low and the biodiversity is high. However, because of human activity, sometimes fertilizers come into the lake and after that phytoplankton can experience a bloom and consume all the oxygen, which will kill the fauna. So we have we transition into this undesired state that in which the biodiversity is very low and the nutrient concentration in this case is very high. So again, this is also an example of a stable state and the system shows some hysteresis here because once you once you reach this stable state is very difficult to come back to the previous one, even if you stop throwing more fertilizers into the lake. And there are many other examples such as in our coral reefs that are challenged by the increases in temperature and pH of the environment. And in ecology, many of much of the work has been done with a focus on microorganisms when we are talking about the effects of climate change. But given that microbial communities are at the very heart of many ecological functions and can support biodiversity. It is becoming more and more clear that it's necessary to also study the effects of these changes in the environment on microbial communities in different in our different ecosystems. So my research focus on studying how microbial communities respond to perturbations. And before I go into the specifics of the talk, I will just give you in a nutshell a scheme of how I usually approach this this my research. So natural ecosystems have a very high complexity. And usually this is correlated also with a very low controllability when you want to do experiments. And on the opposite side of a natural ecosystem, maybe the most simple ecosystem in which you have a lot of controllability that you can think about must be a test tube in which you put one or just a few species. And when you do these, in many cases, you are able to to study the interaction network between the species how they interact also with their environment. And, and you can really have a powerful, you can reach good predictions about what is driving the dynamics of the system. So this is how I usually start my experiments. And then once I learn about the simple system I increase the complexity of the network of interactions within the microbes, or even also in the environment. For example, more recently, I've been studying microbial communities inside a more the gut of a worm, which makes the environment to be more dynamic and more complex. So all of this research I do it in in constant feedback with theoretically driven hypothesis. So I will start with a simple model that tries to capture a specific phenomenon. Then I design an experiment to see if I can, I can observe this and depending on my observation. In many cases, I have to rethink my model because because it obviously you don't get it right at the first time. And so in. So, in order to apply these these approach to our research we see which is the usual approach that we take at the go lab at MIT. My background helped me quite a lot into this because I started my, my PhD in in biological physics with Joachim Ford at University of Girona, and there I was a study of biological invasions by doing reaction diffusion equations model for, for several populations like virus infecting a cell or, or trees expanding into a new territory. This background was very helpful when I joined Rikarsole, Rikarsoles lab, because there I started doing experiments for the, for the first time. It was very fun. I liked it a lot and we started to manipulate engineer some bacteria and manipulate their interactions so that we will see also what impact these interactions can have on the special patterns that they form when they are expanding into free space. So that's, that was just an actual of how I ended up at the go lab. And today I'm going to talk about my, my current research in, in a jet force lab. So for the rest of the talk, I'm going to focus on two research topics. The first one. I'm talking about how communities respond to perturbations. I'm going to focus on a biotic perturbation. So I'm going to alter the composition of the, of the community directly. I, I will show a remarkable case in which an, an invader can enter a community and induce a transition to one, an alternative stable state of the community, even in cases in which the invader is not able to survive this transition. And later I will focus on a different kind of perturbation specifically and a biotic perturbation in which we threatened this community with an antibiotic shock. And in this case, I will, I will be showing some ecological drivers that you can use to control the outcome of this antibiotic shock that might be, that might be threatening different ways, each of the species that we have there. Okay, so I'll begin and as, as I was saying, this is what I've done with Jeff Gore and my first experimental community is going to be made by this by stable pair of species in which I have a corino bacterium and a lactobacillus that are inhibiting each other. We have learned also that the mutual inhibition is given by an antagonistic interaction that these microbes have with the pH. So one of the microbes is increasing the pH, sorry, decreasing the pH which means increasing the proton concentration in the system. And when it does that, it's also benefiting from this change in the environment, while the other tries to increase the pH because a higher pH is beneficial for itself. So we can also see this when we culture these two microbes in isolation. So if you have a test tube with only the green species, you will see that the pH goes up and then it stays up. And these, and this helps this microbe to grow and, and live in good conditions. If you start with the orange microbe, you will see the pH to go down and the microbe is able to keep the pH at a low value. So now we move into the community. So what, what does this interaction do to these, to these microbes when we put them together. So my experiments are usually daily dilution based experiments. So I start by inoculating some microbes at a low density in fresh medium when, and then they can consume the nutrients and grow. And then grow to saturation and the next day I take a sample and I put them into fresh medium again. So we go into the cycles of growth and saturation and growth and saturation. So for this community, what I see is that when I start with a relatively high abundance of the orange microbe. You can overcompete is the green microbe and then take over the population. So the main interaction that you find here is the orange inhibiting the green. However, at a low initial relative abundance of the orange, we will see that its abundance goes down because the green microbe is taking over the system and therefore the dominating interaction is the green one inhibiting the orange one. So this by stability is actually robust to some degree of migration. So even if I put a low amount of fresh cells entering the system every day, the, the two stable states of the system do not change. In other words, I have a by stable community that now I can play with it and perturbing different ways to see how I can obtain transition from one to the other stable state. I was introducing earlier my first perturbation in which I'm going to focus on is going to be an invasion perturbation. So in one of the dilutions. In addition to the migrant cells I'm going to in inoculate a small amount of an invader cell as species that don't belong to this community. And when you do that, there are in ecology to classical outcomes that have been broadly described and one of them is the establishment of the invader. This happens when the invader is able to grow well and compete well with with the local community. And it leads to its establishment. In other words, it happens when the invader is a bad competitor compared to the local community and then it ends up going into extinction. And this is known as community resilience. In most of the cases it's assumed that if a community is resilient to an invader after the invader is extinct, you just return to the previous stable state. However, there's been a largely overlooked scenario in addition to these two, which is the case in which you have an unsuccessful invader. So it doesn't compete well and it's going to towards extinction. But during the time the invader remains in the community. It's affecting the interaction with the other two microbes. So at the end you could have a transition to the alternative stable state of the system. We have called the transit invader. And in now within our experiments about invasion perturbations, we were wondering whether we will find such a case. And actually we did. And it was very exciting for us. Let me show you this experiment in which we start with the orange micro dominate in the community in a stable way. We introduced pseudomonas chlorophyse as an invader, and we can see how pseudomonas goes up during the first hours. And by doing that, we also see an increase in the population of the green micro. And after just two cycles. Pseudomonas is reaching the extinction, but in, but the competition between the local microbes have totally changed. And at the end of the experiments we have observed a transition into the alternative stable state of the system. And now the population of the orange micro is very low. It means that we can see this invader as a perturbation that drive the system from one of its stable states to an alternative one, even if the invader did not survive. So here we wanted also to know about what will be the mechanisms driving this transition. At the beginning, I've told you that the orange micro have an interaction with the pH in which it's trying to increase the pH and benefiting from that. The opposite is true for the green micro which tries to, to, to, sorry, to increase the pH. I'm not sure if I said it well before, but basically this one increases proton concentration this one decreases proton concentration. And the interaction that the invader has with the pH is different. So this micro will benefit from a low pH. However, its metabolism is increasing the pH. And when we measure the pH during the invader experiment, what we observed is that the orange micro was keeping a low pH in the system. And right after we introduce the invader, we saw a dramatic shift in towards a high pH. And once it high pH, the green one was a better competitor than either of the other two. So here we have a mechanistic explanation for what happened to our simple experimental community. And our next question was, could we observe this in a more complex community. And driven by, by this motivation. I started a completely new set of experiments. And in here, what I, what I did is to take a soil sample from the yard that it's just in front of our lab. And I transferred this sample into our test tubes. And again, continue with this daily dilution. So even if there are many different species coming from the soil and I don't have too much control over them. When I did this experiment, I observed that for many different replicates, each line here, it's a different replicate. The community will go through an initial transient phase of some pH fluctuations, but eventually many communities, not all of them. Here are more or less 50% of the ones that I tried. And the other 50 were not stable at the end of this amount of cycles. But for this 50%, I could see how it looks like after a few days they have reached some stability. And when I looked at the composition of them, I could also observe that for each of the different pHs, I was obtaining a different community there. I'm going to focus now on this one in cream color and blue color here, because for this one, I, I performed the next experiment. I'm going to try to see if they are also stable states of the system by introducing some amount of migration between them. So I did this and I observed that even if in my daily dilutions, some of the cream microbes were coming into the blue community, the blue community didn't shift to the alternative stable states and vice versa. However, in the same conditions, I applied an invader to the, to the community with a lower pH. And it's the same invader as before with my experimental community. And what we can see is that after doing the course of a few cycles, the pH goes up and it stays in the range that is characteristic of the cream community. So not only this, but we sequence the community at the 16th level. And then we could, we could observe what was the composition of this community. And indeed it was, it was dominated by the bacillus genus during the first days. And then I introduced the invader. And this pseudomonas goes up and down. And after this, we can see how we have transitioned to a to an alternative community that it's in this case, governed by Pantoea. So here, again, I have this invader induces a transition to Anna towards an alternative stable state. And it's a good moment to summarize my first part of this talk. So I've been showing that environmentally mediated interactions are important drivers of community dynamics. Sorry. Yeah, yeah. There is a question by Matteo. Yeah, so. Oh, okay. I'm sorry. My speakers go off if there's no sound coming in a while. So I'm not sure how much time have you been. No, I wanted to ask a question first. Yes, I was about to say that it's a good moment for questions. So please go ahead. Yeah, so the last example are the communities homogeneous in the sense that are they dominated by a single species. Yes, that's that seems to be right. So, yes, so when I look at the composition of these communities in terms of colony morphologies and also where the different strains lead the pH for the blue community there's some heterogeneity. So I can see three different kinds of morphologies in in these bacillus strains. And some of them are more associated to a pH that it's around six. And the other ones more around four and a half or so. So there is some heterogeneity in this community. However, they are very close relatives because the 16 s sequence ended up being the same for all of them. And for the other micro as well so Panto is largely dominating this community. And I only have very, very low fractions of other species that include all of the other species that were in the soil community but they are, they are kept at very low levels. I have another question. Yeah, at time zero when you sort of have this invasion of pseudomonas. Do you also sort of put a little bit of Panto air or not. So, yes, yes, exactly. So this was done under this migration condition. So I, if I only do migration, this is the outcome. But if I do migration, in addition to the invader, then I observe this, this transition. Right. So if you don't I mean I guess if you just have the invader and you don't have the migration you are going to see that pseudomonas increase a little bit and then goes down and then you go back to the previous state. So that's what I will expect. I don't remember if I ever did that experiment, but an alternative possibility could be that pseudomonas takes over. Right, because you could think that if there's no Panto air, the pseudomonas is able to overcompete the bacillus and then stay there. So I would say the one that drives Panto air could be the one that drives Panto air. So can I ask a question as long as we're in a question period. This is Simon. We presented an equilibrium view of communities but a standard view of ecological communities is that there's a successional process. Following a disturbance there might be early successional species that will come in and will influence the later development ultimately to be replaced and those species themselves could be initiating the disturbance. So what we're seeing here that the ones that are transforming we can think of as early successional species that are transforming the community. Yeah, I think that's right. That's also a way to to see it. If we will have like if for whatever reason this pseudomonas was not there from the beginning or or it cannot really change the environment in the way that it's supposed to until the bacillus is there. Then we stayed in one state of the community. Then this is as you as you are mentioning Simon, this is a case in which we have a succession and this succession is allowed by this invader that engineers the environment. Thank you. So I was about to summarize. I think the question have helped into that so basically I've shown that these microbes. In general microbes can interact with their environment so that they change the environmental conditions and this has a feedback into the behavior of the micro itself and it's an important driver of community dynamics. As a remarkable example of this I have been showing this case of the transit invader who switches the community from one stable state to another one out surviving the transition. And now I'm going to go into the second, the second part of the of my talk. So, as, as I was introducing earlier invasions is not the only way in which community can ship from one stable state to another. The many other perturbations are also interesting in the case of microbial communities. Antibiotics are especially interesting one for for the obvious biomedical applications. So, before I go into what I did experimentally, let's see about what kind of problems are we facing. So, here, if we think about a complex community, such as the one inhabiting in our, in our gut. So, the, the human gut microbiome. We can see that there are many factors that can influence the state in which the community is. So, here, we can think about the reasonably diverse community that inhabits the gut and, and the diversity and the state of it is going to be determined by different factors such as diet and species coming into the diet and whether you have functional redundancy or not. And then many other host factors such as the age and the environment of the factor of the of the host story. So in these cases to these complex communities in many cases we are applying an antibiotic therapy. And when this happens as a collateral damage from the antibiotic the God suffers and and and we can see a decrease in the diversity of the microbes. So some of these microbes will just go extinct. And others could start dominating the community either during that time or afterward. So in these different phases before doing an antibiotic and after an antibiotic there are all these different factors that can contribute to whether the final community can recover to the previous state. Or it goes into a different one and having a different state in this case can change the function of the community so have an impact on so in the health of the of the host. But, but again given the complexity of the system is difficult to to find a specific mechanisms that that can help us to predict in which cases some patients do not recover and in which cases the patient will recover the previous gut health that they had. So at the heart of this problem is that the research on antibiotics has mainly be driven by studies on single species instead of communities. So when you are trying to test an antibiotic you usually have a pathogen or specific species in mind. What you try is to see if these species is susceptible to the antibiotic. So here is how here's a minimum inhibitory concentration has a that it's the usual tool to assess susceptibility. So what you do is you start with with an inoculum of microbes in test tubes, and you have a gradient of the concentration of the antibiotic. So that you will wait for a little bit of time and you see in which conditions they the microbes growing in which don't. And in this case we have seen that this one here is the minimum in minimum inhibitory concentration so here is the concentrations at which the microbes were already not able to to grow. And the antibiotics the microbes here can just no grow, not grow, or they can also experience death, but this doesn't change you assess the susceptibility of this micro. Even that this is the most usual tool, a classical hypothesis to try to predict what will happen to the community is that the more susceptible species maybe they are going to experience the harshest consequences from from an antibiotic exposure. But of course in ecology, we know that resilience against perturbation is is driven by many different factors. For example, here I'm showing this cartoon in which I have a community with different microbes and and there are signs of by stability, but these these by stability can be shaped by ecological conditions. Ecological factors such as the pH as I was showing before, but also temperature or nutrient availability, even migration so the amount of new cells that enter the community per unit time. And such factors can influence both the shape of the stability landscape and how deep are the valleys of stability, but also resilience which is how strong a perturbation can be and the system will still recover the previous stable state when you remove the perturbation. So in order to assess this interplay between ecological drivers and the susceptibility to antibiotics. I went back to my experimental community. And I started by assessing the susceptibility of my two microbes. So I did this minimum inhibitory concentration, I say, and I observe that the green microbe was more susceptible to the antibiotic. And we can see here that the minimum inhibitory concentration to stop growth of the green microbe is much lower than the than the orange. So based on this, if you make this classical hypothesis, you will see Oh, well, maybe then the orange microbe is going to display the more resilient state. So I started then. Yes. In this case, when you say resilient, you mean what is the basic of attraction. So depending on the initial condition. I mean, how wide is the set of initial condition for which you go to the tractor with the orange species, or you mean how fast you go back to the tractor. I mean, yeah, in this case is for now it's going to be more like a binary thing, whether the stable state will return to the original composition or not. So, actually, I'm going to, I'm going to be talking about relatively hard or strong antibiotic perturbations, but more like my, my current experiments are studying about the resilience in this in this kind of sense about. Okay, now if I vary the strength of the perturbation of the antibiotic what's going to what's going to happen, but for now I'm just going to make a case for for relatively strong antibiotic perturbation and whether the community is resilient or not in terms of whether we observe a transition or not in the stable states. So we, from here, we were predicting that maybe the stable state dominated by this more susceptible species will be will be less resilient to the to the antibiotic shock but what I observed in doing this experiment in which the community starts in the green in the stable state dominated by the green species. And one day I transfer it into the antibiotic and I see how this is harming the community. Then I remove the antibiotic and I let them recover. I see that the green one comes back and dominates the system again. So in this case the this state is resilient to this antibiotic shock. And when I did the same starting from the alternative stable state dominated by the orange one I applied a perturbation. And when I leave them recover what I see is that the green is taken over and eventually over competing the green micro. So somehow these these goals against a classical hypothesis based on just the susceptibility. And in order to try to understand this better we we began by by starting a simple model that maybe could capture what's happening in our community. In particular, I started with a modified version of the load couple terror model. And here I have two species. I'm I call them F and S because of reasons that will become apparent later and for each species. So species F has a growth rate are F and it experienced logistic growth so the saturation of growth at high densities. And there's also another term for the interaction between the species. So the species S is inhibiting the species F and vice versa. To better capture the dynamics in my experiments we have a constant migration rate which is the same for both species and the death rate that it's going to be applied during for only for a fraction of time. And this will be an antibiotic associated growth rate. So what it means is that the death rate is going to be zero until I apply the antibiotic in which I'm going to have different growth rates. And in particular I'm going to consider that the green one is more susceptible to the grad and the death rate is is higher for this one. And after the antibiotic we go and after the shot we go down to zero death rate again. So if everything else is equal between these two species then the classical hypothesis will work here because you can state with the green one dominating the system and you apply the perturbation which starts killing the microbes. But then after the perturbation the orange one wins the competition against the green one which has been more harsh more harmed by the antibiotic shock. Alternatively we can consider what you will expect in general in microbial communities and it's that you have different growth rates. So now I'm going to consider that F has a faster growth rate than S and if you do that in the same conditions that before then what you can observe is the opposite. So even if you start with the less susceptible species dominating you apply the antibiotic shock which reduces the population abundances. But then after this the green micro can grow faster than the other one and end up dominating the community. So this led me to the question of whether this will be consistent which where what I was observing in my experiment. So is it the green micro really growing faster than the other. And if so we could make a prediction that these should work for many different kinds of antibiotics because it looks like the importance of the growth rates can override the importance of the susceptibility so it's not about one particular antibiotic. So now I went back into the lab and I measure growth rates in several replicates and actually we observe that the green micro was a faster grower so this was looking good. But then I also tried many different antibiotics. And in most of the case in most of the cases the green one was always was most more susceptible when when doing this minimum inhibitory concentration assays in isolation. But when I perturb them in communities. I also see that the transitions were usually going towards the green micro with with a few exceptions. So this is an indication that the simple model could be capturing what is happening to this community. So here again is a good moment for questions. This this should be a takeaway message that the first growth increases the resilience against antibiotic shocks in a community context. And if if that was enough, this should be the thing to remember from the second part of the talk. So if you are still with me for five more minutes I will show you how migration can also tune these growth rates and hence the the antibiotic response. Is there any questions so far. Okay. So, when I was measuring the growth rates I also observed that my that the green micro was a beating some signatures of cooperative growth. And that is that for different initial cell densities, I will observe a different growth rate for the green micro while the orange micro was mostly growing at the same growth rate no matter the initial density. In the model this is usually captured by applying an effect which which consists in an additional term into the growth rate of the micro in this case the fast micro the fast grower. And this Ali effect. When we use this form of the Ali effect what we see is that capital growth rate will grow, but then, sorry, go down as you go down in the initial population density before comparison the the orange micro is it's experiencing only logistic growth here. And this happens when we introduce these into our into our theoretical model. So, before we were in the absence of couple of these was my modified set of equations for the lot capital terrible model. And what I'm showing here is how the basing structures are for the for the two stable. Wherever you start in this initial population density range, you will end up with the green one dominating in this in this region of the face space. However, if you if you go into a higher initial abundance of the orange one then you can reach a stable state dominated by the orange. And I'm showing here, it's a, it's an example of a perturbation an antibiotic that it's harming both of the species at the same rate. So depending on whether this is the case or not you will have different inclinations of the arrow. But here, if you are in the stable state you freedom, the microbial community in this way, then you leave them recovery, and it could come back to the initial stable state. If the perturbation is harsher, then there is a risk of crossing the, the interface be two basings of attraction and then is when you can observe the transition towards the green one dominating. So how this changes when you have Ali effects. So if we draw the same phases space here, we will see how these is boundary between the two phases curves in a different way. And now the basing of attraction of the orange microbe is much bigger than the, the one for the green microbe. So we have this prediction that cooperative growth can change the resilience of alternative stable states. And the next question is, could I apply this to observing my system, a change in the direction of these transitions. And the answer could it look like that like the answer could be finding the migration, because migration establishes a floor in the population densities. So let's imagine a harsh perturbation of the antibiotics so only a few cells survive, and then I dilute them into fresh nutrients but I also have some amount of migrants that are going to come and they are going to be responsible as well for the repopulation of the system. If I increase migration, even if the rates of migration that the fraction of the migrants are the same for both the species, we will have a higher initial population density right after the shock. And given the cooperative growth of the green species, it could this could have an effect on the on the difference in growth rates between the two of them. And after doing this analysis I realized that I my migration grade will set up a floor in the population densities that was more or less over here. So the prediction was that maybe if we load a duration rate we will reduce the difference in the initial growth rates after the antibiotics and maybe for many antibiotics we will observe a transition in the opposite direction. Sorry. Yes. So what you're saying is when migration sets the floor of population density and observe only effect I need to be at a small enough population. And therefore I see that effect if I have a low enough migration. Is that what you're saying. Excuse me, I didn't hear well the second part of the question. I mean, what I don't understand is how these. I mean, what is the connection between migration and right. Okay, so imagine. Let's imagine an extreme case in which the the the antibiotic is killing the whole population is that's the extreme. And then what happens the next day is that you don't transfer any cell from here. But there is the migration rate that so some cells will come into the system because of the migration rate. If the total migration rate is low. Then you will have an initial population density that maybe it's around here. So you have this growth rate for one species this growth rate for the other. So if you increase the migration rate a little bit, maybe you are going to start from here, or from here. And this is going to change. I was trying to understand that in terms of the parameters of the model with the effect. So what you're saying is that you should compare the migration rate with a with, which appears in this. Yes, yes, that's right. If the migration rate is more than enough, then you are, you can see the effect of the effect. If it is large enough, you are already above. Yes, yes, you're right. Yeah. Exactly. Yeah, yeah, yeah, the migration rate will set up where you are in this score if you. Yeah. Yeah. Okay, so the last thing to check was whether we could lower the migration rate and and see if this had an effect on the reaction of the community to antibiotics. And just a reminder, when I was in my original migration rate, which now I'm going to call the high migration, great condition. The orange microbe was able to sorry, the green microbe was always able to take over the community after an antibiotic shock, no matter which state I begin with. So high migration means that you have transitions towards the green microbe dominating the system. Now I did the same experiments at a low migration rate. And we actually observed a transition towards the orange microbe. So I begin with the green one. I applied a perturbation. The migration rate is low. So the green microbe is a still a faster grower, but the difference is lower. And we have to take into account also the inhibitor interactions between them. So these allows the microbe the orange microbe to over compete. It's the other species. We also observed that if you start with the orange microbe dominating the system, you end up with the orange dominating the system. So this not only happened for one antibiotic, but for many of them. So when I, when I lower the migration rate and I apply different antibiotics to this community, I will observe transitions predominantly going to the orange microbe. Or in some cases, both the stable states remain resilient to to the to the perturbation. So no changes. But if I apply a higher migration rate, then I'm going preferentially to the green microbe. Sorry, can I ask a question? Yeah, so I see error bars on your plot. So you are averaging on different realizations. So yeah, these experiments. So three, three different experiments for each of the blocks. But in all of three in all of them you observe the same outcome or can you observe also metastable outcome? So for the ones that I'm plotting here, I always observe the same outcome. I do not observe metastable outcomes in the long term. If metastable means that none of them is going to dominate and they are going to exist more or less of the same fraction. What I can observe, depending on the on the strength of the perturbation, because if the perturbation is very weak and you only put a little bit of antibiotic, then this is not going to do anything to the microbe. So you at some range, you don't observe transitions and then you start observing transitions and I'm showing that the migration rate can also reverse the direction of these transitions going to the green or going to the orange. So when you are in experimental parameters that are in between this decision, sometimes I observe that that if I do three replicates, one will go in one direction and two will go in the other direction. Okay. But it's just for very small ranges of parameters. Thanks. And yeah, so this this was all just to summarize, I was showing how environmentally mediated interactions can be an important driver for community dynamics. And then this can lead to remarkable dynamics such as the one from the transit invader who induces a switch from one stable state of the community to an alternative one. And we could, in this case by studying simple communities in the lab, isolate specific mechanisms that can drive this transition. And, and finally, beyond this environmentally mediated. There are also other ecological drivers that we could study more in depth and exploit to control the outcome of microbial communities after perturbation suggest antibiotic shock. And with this, yes, and, and as I was mentioning to Jacob or right now the second part of my talk is ongoing work and what I'm doing right now is going into these regions of parameters in which the different factors are more equilibrated in their impact. So what what happens, the susceptibility is more or less as important as the difference in growth rate, for example. And with this, I would like just to thank you. Thank you, my, thank you, Jeff, which is my, my current advisor, but also my previous advisors, from which I learned a lot from all of them, and my, my colleagues and collaborators Christof who is an author in in the paper about the transient invader and many other friends that are helping me very in my everyday life, like Martina and Jonathan Friedman and many others. That I would like to thank question to answer any questions that you may have. Thank you. So we have time for a few questions. Yeah, I have one question. Yeah, on the first experiment. So you said that the mechanistic interpretation for why the invaders shifts the community to the other stable state is that the invader increases the pH. So I was wondering if you tried just to change the pH in the community without putting the invader and seeing if it has the same effect. Yeah, yeah, we did that in and it has exactly the same effect with with the pH you can control in an intuitive way where you drive the community so low pH you go to one micro dominating and high pH shock and you get the other one. Hello. Hi. Another question is not directly related to your talk but just want to know if there are studies on if you control the new trends, like in a periodic way, that's kind of more like what we do daily know with the food coming in and so on. If we go if we change this hours for example making irregular or even with some fasting period. What would be your prediction, whether it's going to be a significant change in the composition of the microbiota or you know, and maybe it's two general question but I want to know, you know, no, no, it's Yeah, it's a it's a good question. I'm thinking about the impact so my view of the the effects that new has on microbial communities. Usually, I think that a higher nutrient availability increases the competition strength between the species. So when you, when you have a lower availability to nutrients, the microbes not only might grow slower, but also they some of them might be lacking some nutrients. So they relay on on mutualisms. So one micro would be using this chemical that they cannot find in the environment. You increase the nutrients and the nutrient availability and then you remove these dependencies because they just take it from the environment. And now they are just competing to grow and overtake the system. So, instead of changing the concentration if you change the timing of it. Yeah, I think it's going to depend as well as what happens once you reach some saturation and some of them might be secreting toxins that are bad for the others. So it definitely changes the, it definitely changes the composition and the stability. But I wouldn't be able in an abstract way to say in which, in which direction specifically it's going to. But if we have sufficient understanding, it may kind of design some, some schedule, right, for people who like to eat, but then they worry about getting fat and so on. And, and I think that in it can help to do it for particular sets of particular sets of communities or. Okay, thank you. Any other questions. Yeah, so I'm wondering to what extent in such a simple system, one can validate theoretical models. So, so for example, when you can describe this with this as an evolutionary game. You can write down a question so the question is, can you really estimate the utility functions of the different bacteria in this evolutionary game, at least to test whether this paradigm in a simple system is really what is going on. Yeah, so my answer will be that with simple communities, we are reaching the state in which we can really quantify much of, of ecological and evolutionary drivers there. So here I haven't shown it, but, but these, these phases space of the two stable states and the, and the basing of attraction for the two stable space. That's something that you that I have studied experimentally so you can really tune the range of initial concentrations of the two microbes and you really have in one experimental play you will actually see, because you have the different population densities by each micro you really see that there is the, the separate tricks there in your in your actual experimental play. And in, in these cases when you have two or three or four microbes that you have previously studied really well in the lab. The ability that we have right now to ask a quantitative question and go and measure it. That's very high, and the challenge right now is to go into more complex communities so more than, yeah, more than five species or even going into comparing these two to a soil community for example in which the initial number of species is astonishing. Does that answer your question. No, yes, no. I mean, the issue is that if you can validate models in this simple system, maybe then you can think that you can study models with many species. Also from theoretical point of view, I mean, extended. So, for example, with one of the grad students at the core lab right now, we are taking this approach in which we start with complex communities. And the model is based on on statistical physics tools so now you, you start thinking about about the, the community as if you could average the properties of the different species and predict what's going to be the abundance of the average species in the system, right. So this is kind of going a step forward. I'm, we are not able to predict there what happens to one particular species. But when we look at the statistical properties of the community we are finding very promising and and compelling results about how can we apply these different modeling tools to to our communities in the lab. Thanks. Great. Is there any other question. Hi, can I ask a question. Yes, please. I will first introduce myself briefly I'm she and recent graduate from geosciences Princeton University but I will start a post-op position with Alvaro Sanchez, who was in your current lab. And sorry if I crash the seminar today, but it's a very interesting seminar and I wasn't thinking about these problems. So I have a kind of a basic question which is what is, when can you call a microbial community stable. Like, I thought stable community means that if there are some perturbations they can still remains the same, but it seems that you're talking about adding somewhat stronger perturbation so the microbial community community shift. So I was thinking is there a way to, like, is a barrier criteria to tell us what is a stable community. In my experiences. Yeah, thank you for your question in my experiences. There is always a matter of definition. So what do you mean by stability one and resilience and it can change from study to study in my case here. Well, first I, my approach is that if you go into very, very strong perturbations no community are going to be stable. Right. So there's always a limit. So because of that, I understand stability as a relative consistency or persistence of population abundances and ecological functions. And, and the resilience is how much how strong the perturbations. Yeah. The resilience is related to the strength of the perturbation that this the community can endure without without experiencing a change in the in the stable state in which they are. Is that answering your question. Yeah, thank you. So it's more of a relative. Yes, I think it's always going to be a little bit relative, but we can learn once you define and what you mean. And yeah, was you define what you mean by it, then we can learn about about these concepts. I think it's time to thank again, for getting the seminar.