 Mariell Brookner, she is a graduate student at Utrecht University in the Netherlands. She was a runner-up in our student-modeler competition, meaning she had a great submission, and that gets judged both in the sense of the scientific contribution that the students are making as well as their data and code and software development. Her particular topic is the effects of ecology on morphological development of estuaries. So it's very appropriate to the theme of this year's CSDMS meeting. She's also interested in these predictions of morphological change among timescales, but looks at the relevance for like eco-engineering and management. Today she will talk about the modeling, modeling the effect of dynamic salt marsh and microfitobentus growth on the large-scale morphology of estuaries. Thank you, Mariell, you can take it away from here. Great, thanks Irina for the nice introduction. Hello to everyone from Europe, I'm Mariell, and as you can already see in my title today, I don't want of time, unfortunately, to talk about microfitobentus growth modeling, but I didn't strike it out because we didn't do it, but just because of the time issues. So if you're interested in this, don't hesitate to contact me after the presentation. And the background of my PhD and also this presentation is that we would like to understand how morphology of estuaries form and what constrain them, and they are obviously not only driven by the boundary conditions and the constraints by geology or human impact, but there are also a lot of species living in the estuary that affect the morphological development, and this is precisely the topic of my PhD. So we call these species eco-engineering species because they live on top of the sediment or within the sediment and determine the erodibility. And today I will be focusing on the salt marshes. Also on the biostabilizing species. And there are obviously a bunch of different species that can be called biostabilizers. So there's not only salt marshes, but there are also biofilms or, for example, muscle beds. And yesterday we also heard a great presentation about seagrasses, but you can think about many other species that live in the soil and can affect sediment erodibility. But salt marsh presentation is specifically interesting because it directly affects the flow and leads to a de-saturation of the flow velocities, which leads to a sedimentation in the vegetation patch. But what is the most interesting part about this, I think, is that the vegetation is also affected by the sedimentation. But to account for this roughness of the vegetation on the flow, we can account for a roughness coefficient C and for a drag force F. And these two parameters are defined largely by the variables of the vegetation. So how tall is the vegetation? Does the vegetation have leaves? And how dense does vegetation grow? So the effect of vegetation is very much determined by the type of vegetation growing and also the seasonal cycle of vegetation. So as I already mentioned, if we account for the ecoengineering effects, we also need to know how does the vegetation react to the changes that it applies to its environment because it will obviously not grow when flow velocities become too high or inundation periods change. So in order to understand the effect of vegetation on estuaries, we need to capture this feedback loop between the ecoengineering effects on the species on the one hand and the vegetation response. That's why we developed this ecomorphodynamic model that on one hand is a hydromorphodynamic model parameterized in DALF 3D, which is a pretty sophisticated software package that we use a lot in the Netherlands. And this model is coupled to a dynamic vegetation code in MATLAB that accounts for establishment of vegetation growth and mortality. And these two models are coupled in a way that they interact on a bio-weekly basis. So the hydrodynamics that are computed in the DALF 3D model will determine where vegetation can establish and it will also determine if vegetation can survive. So if the flow velocities or the inundation period become too high, then it will kill off the vegetation. And on the other hand, the vegetation will also grow. So because of the growth and the mortality, the drag force and hydraulic roughness on the hydrodynamics will change and this will lead to this feedback loop between the hydromorphodynamic computations and the dynamic vegetation model. So this is not the first model that includes this feedback loop between species and hydromorphodynamics, but what makes this model a bit more special is that it's based on literature, which means that the vegetation parameters are determined by values from experimental studies or field work studies. So we do not prescribe vegetation or calibrate vegetation to a system, but we turn on the system and let the vegetation develop based on the hydromorphodynamic feedback of the model. Especially we also account for aging of vegetation by several life stages. So if the vegetation grows out of the seedling stage, it will become more resilient to stresses and it will also become larger in plant sizes, which will affect the drag force and the roughness again. And in this model, we can combine several species. This is not only restricted to several saltmush species, but also, as you might have thought, micro-fetibendoc species and also a bunch of others. And we already showed in my first publication of my PhD that the saltmush patterns that evolved from this model are pretty well representing realistic saltmush establishment. But obviously, there is not only the biostabilizers that stabilize the sediment, but there is also mud. And this is a nice picture from the Netherlands in case you would like to visit. To be fair, the weather is much better now. But when mud settles in the system, it can create this cohesive cover that significantly reduces the irritability of the sediment and the same as saltmush vegetation kind of sticks the sediment together in place. And the interesting part about combining mud and saltmush is that the mud settling is of course enhanced when there is saltmush present, but there is also the possibility that increasing mud will enhance saltmush establishment. So when we look at a specific site, we don't really know, will we have a mud flat or a saltmush? And this will lead to my research question that I would like to discuss with you today. What does establish first in a system? Do we first get saltmush establishment or does first mud establish and then the saltmush? In order to investigate this question, we take the western-scaled estuary as an example. This is an estuary in the Netherlands and it's characterized as being very dynamic and mostly sandy estuaries, so there's not a lot of mud in the system. And it's heavily dredged and occupied by humans because of a bunch of ports that are established along the estuary. The other good thing about the estuary is because of the ports, it has been studied really well, so there's a lot of data and information available. So there exists a calibrated hydrodynamic model of this estuary in DOF3D. So this is ideal to use to apply my vegetation model to. And this is the DM of the western-scaled and you can already see that there are quite some tidal shoals that can be found in the estuary. And we picked one of them, which is the tidal shoal of Vyde's Orden because it's quite interesting to study because on the right side you see panels with saltmush and mud pattern across time and you can observe that the saltmush establishment took place in the recent decades and has been increasing since and there's also some mud observed on the bar. So in order to tackle our research question, we run a bunch of scenarios and compared them to see to actually answer our research question. And we first start with a reference scenario without any vegetation to see if there is mud settling on the tidal shoal without vegetation present. We have a generic saltmush species parameterized that we run in the model with only sand and then we have a model run where we add the mud to see how important the mud is for the generic saltmush species to establish. And as a last scenario, we have a mud-dependent saltmush species that can only grow in sediment that has a certain amount of mud in the top layer because that's characteristic for some saltmush species that we can find on tidal shoals. And here are already the results. So on the left panel you see the mapping as showed in the previous slide. The dark green colors is the dense vegetation and the light green colors is the sparse vegetation. And then the first two columns are the generic species run. One is with only sand and the second one is with sand and mud through three different years. And you can see that there is not a very big difference between the patterns evolving from only sand and sand and mud in the system, which is interesting because it shows that the mud is not very important for the species to establish on the tidal bar. However, if we look at the mud-dependent species, we see a different pattern. So we have a slight establishment that gradually increases with time and so that the vegetation is still able to establish even though it requires mud in the soil for establishment. We compare the corresponding mud pattern. We again see the ecotope maps on the left-hand side. And the first column is the reference run. So the results of the reference run for the mud in the top layer. The darker the colors, the more mud. And there's already quite some mud settling on the bar without any vegetation present, especially on the southern tip of the bar. So there is mud establishment under certain conditions on the bar. If we look at the two vegetation scenarios, then we see that there is enhanced mud on the bar because there's a lot of mud settling within the vegetation patches. And this pattern is much larger for the generic species than for the mud-dependent species on the right, because the coverage is much larger. But interestingly, even with a species that requires mud to settle, we see more mud on the bar after 12 years of simulation time compared with the reference run, showing that different strategies for soil-mushed establishment can promote also different mud coverage and will enhance mud coverage on the bar. So to conclude and come back to my research question, we saw that depending on the species, vegetation can promote mud accretion in places where no mud can settle because the conditions are too dynamic. So in these locations we first require soil-mushed establishment and then we get mud settling. But there are also other parts on the bar that are calmer and where mud settling can also occur prior to vegetation establishment. And we saw that the vegetation establishment is partly determined by the sediment in the bed, meaning that there has to be some mud to make some vegetation types established and that will alter both the vegetation pattern and the mud pattern. But I've been promising that I will talk about large-scale because this was the title show. So we applied our model to the entire western scale to see if these concepts also apply to the large scale with a larger grid as well. So in the top panel you see the generic soil-mushed cover and on the bottom panel the mud-dependent soil-mushed cover. And we see similar trends with vegetation establishing in similar locations for both species types. But if you look closely at the cover you see that the generic species has a bit more abundant than the mud-dependent species similar to the observations on the title show. And if we look to the corresponding top mud in the western scale we also observe the same locations of mud, but the pattern is a bit different. And to quantify this effect I compared in a bar plot the mean mud area for both of the species types to a reference one. So the blue and the red bar are the increase in mud area with vegetation present. I thought it was interesting to look at thin and thick mud layers which are divided or separated by a thickness of 10 centimeters because thin layers are usually considered to be seasonal and important for ecology while the thick layers are thought to be more morphologically important. And of course the overall mud extent increases with vegetation present but interestingly for the generic species type we have a larger abundance of the thick mud layers compared to the mud-dependent species while the mud-dependent species can promote more thin layer extent because of this gradual expansion. So we can add an additional conclusion point to my slide that also on the large scale the mud area is enhanced when vegetation is present and the way of increase of mud is mediated by the species type. So thank you very much for your attention and I'm happy to hear about any questions. Thank you very much for an interesting talk, Muriel. We'll give people a few seconds, typing seconds, to type in your questions. Andrew asks a question, he says, excellent talk, Muriel. In terms of ecosystem feedbacks, have you considered the role of extracellular polymeric substances, EPSs, in stabilizing mud flats? Yes, so that was precisely the topic that I left out with the microfetabendic species. So we have model runs where we include microfetabendos where we specifically change the critical batchier stress for mud if microfetabendos is present and then looked at how this affects the mud cover in the system. So I'm happy to discuss this further. Thank you. And Abby asks, great talk, did you find that tides impacted vegetation establishment? Yes, interesting. So this model has been calibrated for boundary measurements. So the boundaries that we apply is based on water level measurements. That means that there is also all tidal constituents in storm surges in the boundaries already. But obviously if we would choose boundaries that have a much stronger storm event, we probably get temporarily less vegetation growth on some of the bars. Rose asks this question, she says fantastic talk, can you speak toward conservation implications? Are specific species better to work with for recovery in certain areas of the estuary more than others? I think that depends on the type of system. So I think if we want to conserve a certain area with vegetation type, there's probably species that are more suitable than others. Others might know more about this than me, but I guess that it would be best to use species that are not annual but perennial, so that live more than one year and that are probably also having a bigger appearance. So the morphology of the plant will be larger. So the ecoengineering effects can be larger than with a small species that might have difficulties to survive certain stresses after the conservation project. There's a few questions that are a little bit more technical. So one question from Kim Rong-Ju is nice talk. How do you determine the vegetation induced drag coefficients of different species in your research? So we have this, the equations I showed for the roughness C and the drag coefficient account for the height of the vegetation. And DELF3D calculates the C and the F for all species separately that we implement in the model. And then every species will occur with a certain fraction in the cell. So maybe half of the cell is one species and the other half is the second one. And DELF3D will sum up the C and divided by two. So basically it will calculate a mean value of the different roughness and drag force values per vegetation type. Benjamin is asking a more applied question again too. It's an excellent example of bridging models with ecological data is what he says. But then his question addresses, as you mentioned the Sheld River is heavily dredged. Could your work help to develop more sustainable dredging management and what are the implications for flood control control measurements? I think it might add a certain stone to the big question because I think that vegetation plays quite an important role in retaining sediment in the system. So if we need to dredge and we have less sediment transported into the channel because of the vegetation that will also reduce dredging volume. And the same for flood defense. If there's vegetation in the system then of course this will slow down velocities but that can lead to higher water levels. So I think that vegetation needs to be considered for this question anyways. How my work can contribute? I think that my model can predict vegetation establishment really well. So if we have an area where we want to see if vegetation will establish under certain conditions then we could run scenarios to see if the vegetation establishment will also change flood risk or dredging volume. That would actually be really interesting to try that. I'll give you the last questions by Matt Kearon and he says, do non mud-dependent species also promote mud deposition? And if so, are they specifically promoting conditions that lead to their mortality and replacement? I'm not sure if I understood the question. So the mud-dependent species grows on muddy sediment so by retaining mud in the system it will actually facilitate its survival by the mud settling. So I'm not sure if I understood the question correctly. Sorry Matt.