WEBVTT 00:00:19.030 --> 00:00:26.370 So let us get started. I come back to the second slot um. 00:00:27.120 --> 00:00:36.016 So that was the initial lecture outline. We have covered part one. He and life in moving fluids @unoise@ that was 00:00:36.016 --> 00:00:45.310 designed to be to provide the background to address the second slot, which titled " Fish behavior in response to 00:00:45.310 --> 00:00:46.239 hydrogennamic stimuli. 00:00:46.409 --> 00:00:57.870 Again, the second slot is divided in two lectures. In the first one, we are going to talk about a brown trout and how 00:00:57.870 --> 00:01:09.450 brown try chooses space or habitat. Uh, based on on um, on the hydrogennamic properties of them, of the habitat itself. 00:01:09.459 --> 00:01:15.709 In lecture number two, we will talk about eels downstream migration, um? 00:01:16.329 --> 00:01:23.829 so let us start with the first assessing hydrogennamic space, use of brown trout. Uh. Before I start some 00:01:23.829 --> 00:01:32.074 acknowledgment, all the material that I will be presenting is essentially taken from this paper that was published in a 00:01:32.074 --> 00:01:39.539 journal, Experimental Biology in Ah, twenty sixteen. Ah, I want to acknowledge the contribution of Jim Care, my Phd. 00:01:39.549 --> 00:01:47.711 Student, and Ah, Paul Camp, the other supervisor. Ah, Michael League in Southampton when I was based there. The title of 00:01:47.711 --> 00:01:59.098 " The of the of the paper alliance with the title of the talk. Nothing strange @unoise@ % um what we wanted to do 00:01:59.098 --> 00:02:10.412 here in this study was getting a little bit about their understanding on how, ah, not not much about how ah fish swims 00:02:10.412 --> 00:02:19.758 while migrating, but actually how fish, when not migrating, chooses his space, is habitat as a function of only 00:02:19.758 --> 00:02:29.303 hydrogennamic properties of this avid . We wanted to isolate the effect of on space use that is what it is 00:02:29.303 --> 00:02:36.779 called and um, this clearly has, uh, lots of important implications. Uh, understanding to see how physical environment 00:02:36.779 --> 00:02:44.568 influenced distribution and movement of fishes is is a fundamental team in equatoris. We need to understand what our 00:02:44.568 --> 00:02:52.683 suitable habitats for fish under, you know, ah, ah, as a as a function of many variables for the validity, atrogenity 00:02:52.683 --> 00:03:01.447 and blah, blah, blah, everything that Ah, that describes the the college of the system. But what is the role of 00:03:01.447 --> 00:03:12.079 hydrogennamics? what is really the driving principles that moves fish towards one place or another? how do they choose a 00:03:12.079 --> 00:03:20.287 where to stay in where to hold station, or where they spend their time essentially in aquatic environment. 00:03:20.287 --> 00:03:31.387 @unoise@ % um. This topic has created a fairly big debate in the literature. Um? because if you want to associate, uh, 00:03:31.387 --> 00:03:40.834 what hydrogennamics. Sorry. Well, if you want to, uh, identify what sort of uh hydrogennamics fish prefers. You have to 00:03:40.834 --> 00:03:50.493 find an matrix to describe it. Is it fish? do they like low cost areas to stay there? do they like intensity 00:03:50.493 --> 00:03:58.239 do they like um certain values of the schooners of the flatness of the last year, of any statistics that you 00:03:58.239 --> 00:04:06.868 can think of. And there are, and there is a. There is a huge number you can use and choosing the appropriate rhythmic 00:04:06.868 --> 00:04:15.336 matrix to describe the habitat and the suitability of the habitat again, has been the subject of controversy. And 00:04:15.336 --> 00:04:23.645 reason is the traditional methods. Essentially, when I say that there is a controversy. I'm saying that there are 00:04:23.645 --> 00:04:31.591 conflicting results in the literature. For example, some people say," Ah, fish like to stay in Ah, in Ah, highly 00:04:31.591 --> 00:04:40.226 turbulent waters. Some others say no. They like to stay in count waters. Why are they looking for food? are they 00:04:40.226 --> 00:04:43.269 resting what what? what is what is going on? 00:04:43.560 --> 00:04:55.273 and one problem when, uh, those studies that try to isolate the fact of on the species um tend to heavily rely on 00:04:55.273 --> 00:05:04.228 simple correlation between a single point time I reached the lost statistics and space used by fish. So traditionally, 00:05:04.228 --> 00:05:12.015 what you do, you do experiments in a flum, you put some sort of uh rocks, or or sources of complexity for the hydro 00:05:12.015 --> 00:05:20.709 dynamics. Then you put the fish inside the flu, and you monitor with the camera where the fish goes, you see what sort 00:05:20.709 --> 00:05:28.309 of preference they have for different Ah for different places in correlate, where the fish stays longer with the and 00:05:28.309 --> 00:05:35.539 hydrogennamic matrix of of that place. And what they do is they use time out at single point. The lost statistics. 00:05:35.550 --> 00:05:42.218 However, as we have learned in the previous slides a single point, so that the characterization of the velocity 00:05:42.218 --> 00:05:48.536 statistic at one point does not capture the face to capture flowfish interaction whereby where flow 00:05:48.536 --> 00:05:56.105 properties varies in time and space, simply because it is a one point description. So they would fail to capture 00:05:56.105 --> 00:06:03.284 for example, where you have movement of fish, Aslam, uh, daggering among vortices that would be very difficult to 00:06:03.284 --> 00:06:10.954 capture with one point measure or entertainment? well, it is a whole flu field that, uh, that really describes why the 00:06:10.954 --> 00:06:16.439 fish decide to and train around the cylinder. So that is a shortcoming of traditional methods. 00:06:17.209 --> 00:06:26.589 The other problem is that attempt to quantify hydrogennamic space use by fish was addressed in field experiments, and is 00:06:26.589 --> 00:06:35.031 obviously the mixed results and again, this is not surprising, because in the field field experiments are 00:06:35.031 --> 00:06:43.433 extremely complicated. There are many confounding variables that are impossible to control, like food, presents of 00:06:43.433 --> 00:06:52.378 predators, competitors, mating issues and all these variables will will will hide the real effects will make impossible 00:06:52.378 --> 00:07:04.306 to isolate the facts of hydrogennamics. The third reasons, and is that, ah, ah, previous papers tend to again find 00:07:04.306 --> 00:07:13.881 statistical links between spaces and turban, flow characteristics, balanced statistics that have no clear biological 00:07:13.881 --> 00:07:22.989 meaning. So some people say," Okay, I've noticed that there is a strong correlation between space use and the turbulent, 00:07:22.989 --> 00:07:23.879 sheer stress. 00:07:25.600 --> 00:07:36.629 Why? why would the fish stay in in a region? why sheer what is the reason for tal was the bio bio mechanics behind that. 00:07:36.639 --> 00:07:46.129 So it was sort of randomly selected, was just an observation without a clear biological link. 00:07:47.019 --> 00:07:56.262 Um, so I decided to take a very easy approach, in my opinion, where they made a couple of hypotheses, uh, which are 00:07:56.262 --> 00:08:03.499 fairly intuitive to me. Um, hydrogennamic spaces is dictated by energy conservation. So essentially, this is a general 00:08:03.499 --> 00:08:13.237 principle. If mm fish can save energy has energy, then to do other things that are important for his life finding 00:08:13.237 --> 00:08:17.230 praise or escape from predators, mating and so forth. 00:08:17.600 --> 00:08:26.914 And more specifically, the the choices are dictated by strategies that minimize the cost of 00:08:26.914 --> 00:08:36.719 swimming that as we have seen, have two sources, thrust and and dragon, which are really, really difficult to quantify. 00:08:36.919 --> 00:08:45.613 For for a fish that swims in a terrible and flows in a complex environment. But anyway, this is the first hypothesis we 00:08:45.613 --> 00:08:59.019 did. The second hypothesis was that in an complex environment, we can explain space use by means of two things. 00:08:59.029 --> 00:09:07.200 Essentially, one is the specialized behaviors that we have observed earlier on, so on, and training, entertainment. 00:09:07.210 --> 00:09:16.721 Sorry, a bow wake riding these are all strategy or specialized behaviors that allowed to save energy. And then we 00:09:16.721 --> 00:09:25.852 have built an adult drug function that whenever these specialized behaviors are not feasible or not possible, allows, 00:09:25.852 --> 00:09:33.485 uh, to quantify, Uh, energy expenditure essentially okay. @unoise@ this drug function D. There is nothing new @unoise@ 00:09:33.485 --> 00:09:42.508 perhaps it is a bit new in the in the world of fish, but not definitely not but the way we have built it 00:09:42.508 --> 00:09:51.063 goes as following. Following a simplified example we say let us consider just an object of a of a standard shape, 00:09:51.063 --> 00:09:59.867 like a cylinder or a disk in a 2D problem in the absence of turbulence. Will you put an object in a current, the drug 00:09:59.867 --> 00:10:05.925 exerted on the object is proportional to the square of the velocity. That is what we say with a fish. The dragonfish 00:10:05.925 --> 00:10:12.542 depends only nearly on the lost. Okay, we are in the inertia region. We are far away from there, from the Lorraine of 00:10:12.542 --> 00:10:21.163 snub. So that is that was the first hypothesis and other hypothesis. If we had turbulence as like a fluctuating vertical 00:10:21.163 --> 00:10:30.170 or lateral component, in this case @unoise@ um. The drug @unoise@ ah increases okay @unoise@ % uh without 00:10:30.170 --> 00:10:38.673 turbulence would be it would be um, eh? related to the square of developments. Whenever you are in an 00:10:38.673 --> 00:10:45.799 idealized case @unoise@ a fluctuating latter. I component the drug @unoise@ if. You do the math becomes proportional to 00:10:45.799 --> 00:10:53.630 the velocity multiplied by the square root of the salmon of the squares of the two velocity components, you and V. This 00:10:53.630 --> 00:11:01.454 is very easy to show, and this is due to the fact of the nonlinearity dependence between drug and and velocity. The 00:11:01.454 --> 00:11:10.541 reason why we wanted to put some, some turbulence in this hard dog function is the turbulence, in our opinion, 00:11:10.541 --> 00:11:18.235 destabilized fish. I mean, there is quite a bit evidence of or on that @unoise@ um. Despite the result prevalent in 00:11:18.235 --> 00:11:25.927 order that that we have seen, but clearly it must have um, eh? it should have an effect @unoise@ if not well. We, the 00:11:25.927 --> 00:11:33.764 the idea is to test these apologies with the experiments @unoise@ but. We thought this was the best candidate to 00:11:33.764 --> 00:11:40.984 describe energy consumption by fish while swimming in a hydrogennamically complex environment. So eventually we decided 00:11:40.984 --> 00:11:51.430 to," Oh, let me let me comment a little bit more. This drug function essentially what this drug function says is that 00:11:51.430 --> 00:12:01.564 imagine that you are. You are riding your bike. Okay, there is no wind. You are riding your bike that allows to you, you 00:12:01.564 --> 00:12:09.958 receive, you feel a drug force that is proportional to you square. If you have a latter right component. So some wind, 00:12:09.958 --> 00:12:17.159 for example, starts blowing from the side is going to be more tiring for you. You probably have experienced that. And 00:12:17.159 --> 00:12:24.147 that is because of the nonlinearity between drug and and you. That is what it says. If you imagine that these velocities 00:12:24.147 --> 00:12:24.840 change rapidly. 00:12:24.990 --> 00:12:33.533 You can define an a distract function as you multiplied by the Rutman Square of the sum of the of the square of the 00:12:33.533 --> 00:12:41.595 velocities, any sort of capture that clearly in a turbulent field. You don't have a constant latal velocity, or even a 00:12:41.595 --> 00:12:48.856 constant velocity changing sign. You have that we decided to take as characteristic velocity of turbulence. The standard 00:12:48.856 --> 00:12:59.356 deviation of the loss is at each point clearly, in experiments we are in a 3D condition. So we have to consider 00:12:59.356 --> 00:13:09.310 the lateral and the vertical component to test these hypotheses. We decided to do some fluke experiments at the 00:13:09.310 --> 00:13:18.357 University of Southampton. There is this large flu which is well suited to do some fish research. The flu is long, 00:13:18.357 --> 00:13:27.245 eighteen meters, I think, and rough flee one point four meters, not exactly one point five, but fairly large @unoise@ 00:13:27.245 --> 00:13:39.546 um. We have selected a test section of the flu @unoise@ roughly, three meters long which, was bound by upstream by 00:13:39.546 --> 00:13:50.839 a honeycomb sort of a flow straightening device that we want to have to minimize any fictitious turbulent coming from 00:13:50.839 --> 00:14:01.135 upstream, which is unfortunately in that flum is is fairly high. Then we play in order to create some sort of 00:14:01.135 --> 00:14:09.173 hydrogennamic complexity. We have placed some cylinders of wearing diameter from nine centimeters down to one 00:14:09.173 --> 00:14:16.644 centimeter, which were placed in two different arrangements, increasing. And with this arrangement identified by 00:14:16.644 --> 00:14:25.159 treatment as treatment being this figure. This the idea was to create an editor, dynamic, complex environment with some 00:14:25.159 --> 00:14:35.644 sort of species, if it had the size down, well generated by in the wake of the cylinders um on the downstream part of 00:14:35.644 --> 00:14:44.632 the test section, we put the mesh screen. Uh, with a released chamber for fish, meaning that before each experiment, we 00:14:44.632 --> 00:14:55.175 would put the fish in here wait for a while so that the fish would climbatize and then we would release a gate, open a 00:14:55.175 --> 00:15:02.766 gate in here, so the fish could start exploring these complex environment and choose positions he liked, 00:15:02.766 --> 00:15:11.622 essentially, so that we could correlate, then the properties of that place with Ah, with the choice of the fish, 00:15:11.622 --> 00:15:18.802 essentially besides treatment, a with the cylinders in this configuration treatment B with this other configuration. We 00:15:18.802 --> 00:15:26.647 also did what the biologists call a control experiment so the same experiment, exactly the same thing without 00:15:26.647 --> 00:15:29.549 cylinders. So very, very, very simple. 00:15:29.789 --> 00:15:36.170 And we repeated that @unoise@ um. And I think that is it. 00:15:36.789 --> 00:15:47.920 What sort of fish did we use brown trout again, the source. We got some hatchery fish, and we have divided it in a 00:15:47.920 --> 00:15:56.629 small, medium and large size class, ranging from with a length of fourteen centimeters up to twenty, eight centimeters 00:15:56.629 --> 00:16:06.045 for the large trout case quite a few number of experiments. This is very time consuming and and very difficult 00:16:06.045 --> 00:16:16.027 experiments to do um fifty three trouts for the small twenty five or medium, let us say, in twenty six, for the 00:16:16.027 --> 00:16:25.669 large a trout. We also catch some wild trouts in the field, which were mixed in size were like nineteen centimeters, 00:16:25.669 --> 00:16:35.090 plus minus three point six. Essentially, we wanted to have both hatchery trouts and wild trout to see % uh. If the 00:16:35.090 --> 00:16:43.074 results were were actually, if the hatchery results were lining with wild trout results that any result we got was just 00:16:43.074 --> 00:16:51.059 a result of being grown in a hatch because what we are interested in are actually wildfish but obviously 00:16:51.059 --> 00:16:57.503 there are regulations. You cannot fish many and are difficult to fish. It takes time. They die when you transport them 00:16:57.503 --> 00:17:05.030 to the lab. I mean, it is, uh, you need to have the proper facilities to maintain them in a hydraulic slab. It is a lot, 00:17:05.030 --> 00:17:13.768 a lot of work. Um, all the experiments Eh were done by placing a camera on top of the of the fish, putting some infrared 00:17:13.768 --> 00:17:21.995 illumination. All the experiments were done at night. So zero visibility, the the fish would uh choose their uh, 00:17:21.995 --> 00:17:32.704 their position just by sensing the flow by means of the system. The latter line, we didn't want to have any side 00:17:32.704 --> 00:17:42.329 effects, essentially um so that we could, we could monitor its position. Uh. We also, uh, not exactly automatically, but 00:17:42.329 --> 00:17:44.699 as automatically as we could. 00:17:45.329 --> 00:17:54.794 We did careful measurement of of the floor field with an Av, which is not @unoise@ Ah. Let us say @unoise@ particularly 00:17:54.794 --> 00:18:05.659 a great for turbulence research, but it um it. Ah, it works fine for this type of studies. Ah it is a it is a ah 00:18:05.659 --> 00:18:13.383 allows to make one point measurements by sampling time series of in three the all the three components 00:18:13.383 --> 00:18:20.206 of the of the velocity vector. Frequencies vary between fifty and a hundred Herz. But that doesn't really matter, 00:18:20.206 --> 00:18:27.574 because the sampling volume is fairly large. So the the measurements are filter it, but I mean, ed Are Ok to 00:18:27.574 --> 00:18:33.890 to measure the bike properties of flow, like up to the second order of the statistic. I think we are fine. 00:18:34.019 --> 00:18:43.337 So we could plot the floor field for the mean velocity, the turbulence intensity, which is essentially the variance of 00:18:43.337 --> 00:18:51.354 the velocities, the relative turbines intensities for any component. The Turban kinetic energy, the sheer stress 00:18:51.354 --> 00:19:00.817 the the Reynolds um sheer stress, which is the between the longitudian and the latter velocity. Please keep in 00:19:00.817 --> 00:19:08.920 mind that in this configuration @unoise@ um. Ah, the the the wake of the cylinders would have just one artistic 00:19:08.920 --> 00:19:17.366 components of would be add this willy will mainly with one direction. All right, it is a fairly 2D problem, which 00:19:17.366 --> 00:19:27.756 allowed us to monitor the fluffy lonely at one plane and to make things relatively easy. Um, I have. Well, we have 00:19:27.756 --> 00:19:35.131 chosen to um well to to address these turbulence methods, because these are those that are under investigation, or where 00:19:35.131 --> 00:19:41.769 under investigation were investigating in the past, in the past, literature @unoise@ so saying the turbans are stress. 00:19:41.779 --> 00:19:51.313 This is the the size that we have, that we have estimating, actually using two point correlations. We didn't trust the 00:19:51.313 --> 00:20:00.859 tailor hypothesis in within the near wake zone. So we could actually estimate the the the size of the Ed is by, by 00:20:00.859 --> 00:20:04.519 means, uh, a little more, a little bit more carefully. 00:20:04.619 --> 00:20:13.614 And as you can see, it is a fairly rich in hydrogennamics domain. You have different the sizes you have Ah @unoise@ 00:20:13.614 --> 00:20:18.627 regions with the high robust is high turbulence intensity so high @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ 00:20:18.627 --> 00:20:22.461 @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ 00:20:22.461 --> 00:20:26.294 @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ 00:20:26.294 --> 00:20:30.128 @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ 00:20:30.128 --> 00:20:33.962 @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ 00:20:33.962 --> 00:20:37.795 @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ 00:20:37.795 --> 00:20:41.629 @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ 00:20:41.629 --> 00:20:45.463 @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ @unoise@ 00:20:45.463 --> 00:20:52.540 @unoise@ actually also an effect as we will see a slight effect on on the space use used by the trial 00:20:52.540 --> 00:21:01.183 @unoise@ um in doing um this sort of experiments, um? we had to find a way of quantifying space preference. Okay, we 00:21:01.183 --> 00:21:11.252 wanted to build up. Um, uh, a way to quantify what fish prefers. And this is not that straightforward, because while 00:21:11.252 --> 00:21:21.984 moving fish none of the fish moved enough to experience the whole flowfield so we did decide to define the 00:21:21.984 --> 00:21:22.960 space used. 00:21:23.299 --> 00:21:31.323 And this is identified by these blue points that that are essentially the snout position in time of the fish so 00:21:31.323 --> 00:21:35.699 that is where the head of the fish is a sensor, essentially. 00:21:36.710 --> 00:21:43.974 And then we have identified the sampled. So what they acquired actually, what the fish experienced while navigating in 00:21:43.974 --> 00:21:52.975 the test section. And that depends, as we have seen by the mechanical system they have by the latter line, which 00:21:52.975 --> 00:22:01.463 accordingly, according to literature, allows to have an idea of the hydrogennamics within this range. So, eh, within a 00:22:01.463 --> 00:22:14.652 , the field of detection is, say, two body lanes long and one body length wide around the cylinder so that we could 00:22:14.652 --> 00:22:25.548 map the sampled hydrogennamics and the used either dynamics, higher dynamic preference how did we define it 00:22:25.548 --> 00:22:36.927 essentially, what we did by monitoring the space used we decided to take an either dynamic matrix. Here I have 00:22:36.927 --> 00:22:45.680 taken, by example, the drug function that we like a lot. We say, okay, this is the fish, the space that was used by the 00:22:45.680 --> 00:22:49.879 fish. And we have reported that in a frequency distribution, a plot. 00:22:50.160 --> 00:23:02.453 Ah, um. Then we did exactly the same. Ah, ah, about how this this drug function was % um was common in within the space 00:23:02.453 --> 00:23:12.135 sampled hydrogennamic preference was just defined as the ratio between space used and space and space sample just to 00:23:12.135 --> 00:23:22.610 reflect the fact that we can only define Ah, the preference, the fish % um. The show showed within the rhythmic 00:23:22.610 --> 00:23:34.866 environment. He managed to um to to sample to feel okay. That was that. That was the the." the idea, um farthermore, uh, 00:23:34.866 --> 00:23:47.599 with um, identified all those areas within the the test section, where specialized behavior zones could actually occur. 00:23:48.200 --> 00:23:55.702 If you remember we talked about carmenating, which was known from the literature. So by analyzing the data from 00:23:55.702 --> 00:24:03.646 the numerous papers on we could get a sort of a definition of an air. Where could Carmen get it could occur, which 00:24:03.646 --> 00:24:08.610 depends on the cylinder diameter and the and the size of the fish as well. 00:24:08.990 --> 00:24:17.802 So we could identify it in our test section. The green rectangles identify where fish would go to do and trainment 00:24:17.802 --> 00:24:23.140 again. We could get that from the literature bow riding again. 00:24:24.150 --> 00:24:32.496 There is fairly robust literature that allowed to find potential areas where fish could do bow riding. We also observed 00:24:32.496 --> 00:24:40.693 some new um um sort of strategies. One is taale. Holding is pretty trivial. Essentially, if some fish were very, very 00:24:40.693 --> 00:24:48.502 lazy, and they would put their tail against the downstream a screen and their head against the latter wall, and they 00:24:48.502 --> 00:24:57.056 would stay there without moving so that is not about it and now occurred really that there is a lot of 00:24:57.056 --> 00:25:04.128 clustering of points here and here, identified by the black rectangles, the other one, which is less straightforward. We 00:25:04.128 --> 00:25:08.180 notice a lot of fish staying in this in this area. 00:25:08.630 --> 00:25:16.965 And to be perfectly honest, we don't have an explanation why they really liked that position, because that is actually, 00:25:16.965 --> 00:25:25.340 again, an error of very fast velocities @unoise@ Ah, which would suggest high drug conditions, right? so um high 00:25:25.340 --> 00:25:26.620 energy expenditure locations. 00:25:27.059 --> 00:25:37.674 However, it is not that far off from areas where the fish wouldn't train so we have we have interpreted as a 00:25:37.674 --> 00:25:48.387 sort of but that why there is energy saving in proximity of a lateral wall and a cylinder, eh? we are not that 00:25:48.387 --> 00:26:00.148 sure. Okay, but again, these are the potential areas at low energy cost. Okay, for the fish again. So just 00:26:00.148 --> 00:26:10.207 to summarize we have complex environments we could class five with with the last measurement we had put fish 00:26:10.207 --> 00:26:20.020 inside. And we have observed fish more movements within this test section. And we now have a tool to quantify preference 00:26:20.020 --> 00:26:30.244 expressed by the fish to the different Er. Locations and this is these are the results we got for the drug 00:26:30.244 --> 00:26:41.999 function. Now here these curves on the left vertical axis we have the preference has defined previously B, H and Um. 00:26:42.009 --> 00:26:52.251 Different lines, Uh, are associated with trouts having different dimensions. Small, medium, large and wild class. Okay, 00:26:52.251 --> 00:27:02.657 the estagrams, the gray instagrams instead are um. The frequency distribution of the whole available hydrogennamics in 00:27:02.657 --> 00:27:11.595 the flow domain. Ok eh meaning if these curves followed, the preferencecers followed the the these patterns 00:27:11.595 --> 00:27:19.686 of the Eastern. That means that they, they didn't really like it. They they stayed there simply because that is 00:27:19.686 --> 00:27:27.251 the most um probable hydrogennamic metric. Well, the the in this case is the is the most probable drag that you 00:27:27.251 --> 00:27:33.706 can have. So it is more likely to experience it, because it is everywhere. Okay? if these these curves diverge from 00:27:33.706 --> 00:27:37.439 these maximum, then there is a preference which is meaningful essentially. 00:27:38.910 --> 00:27:49.865 Now, let us forget a little bit about a figure B. Let us look at Figure A. When we did, we computed the preference for 00:27:49.865 --> 00:27:59.749 small, medium, large and wild fish using the function that we defined. We noticed that there was a very clean 00:27:59.749 --> 00:28:06.994 results of a of a biomodal distribution, essentially where they were strong preference for really low drug areas, 00:28:06.994 --> 00:28:15.318 definitely lower than than than than than they are available. The most problem are available at the dynamics, and but 00:28:15.318 --> 00:28:25.811 also a peak at very high drug. And this was surprising at the beginning. But then we decided to filter out all the 00:28:25.811 --> 00:28:35.490 points that belonged to the Um, to these S, B, Z, the specialized behavioral zones. So we took out all the points 00:28:35.490 --> 00:28:43.161 belonging to these areas, and essentially what you get is that this pig disappears so. This preference in this 00:28:43.161 --> 00:28:53.172 drug at this high drug are explained by the fact that the fish was at locations where he had to use, not, uh, well, he 00:28:53.172 --> 00:29:02.613 he was using either or entertainment which is, uh, associate with very fast philosophy. Okay, but yet it was 00:29:02.613 --> 00:29:13.615 experiencing a low energy consumption because of the specialized behaviors we talked earlier on. So this picture is that 00:29:13.615 --> 00:29:25.429 suggests that wherever this as beset do not occur. Fish like to stay at at low drug and to compare now the 00:29:25.429 --> 00:29:34.030 performance of the drug function. Notice this. This result just says we were right. The black function works, explains 00:29:34.030 --> 00:29:38.409 and nicely the space used, the dynamic spaces of brown trout. 00:29:38.579 --> 00:29:48.644 So what we did, we then filter out all the all these points of, as be said, neglecting the behavior in this. In these 00:29:48.644 --> 00:29:56.866 specialized zones. And we compare the performance of the drug function against many other matrix um in this plot, you 00:29:56.866 --> 00:30:05.471 have the results of treatment A with the cylinders, Ah, decreasing Ah from left, right, the treatment B, where they were 00:30:05.471 --> 00:30:12.910 in an were put in another Ah configuration. And the control experiments essentially an experiment with no 00:30:12.910 --> 00:30:23.333 cylinders in the test section on top. Here we have the the preference curves for small, medium, large and Wildf, which 00:30:23.333 --> 00:30:32.665 are the colored curves again. Ah and Ah, at the the the metric serious drug. And again, there is the gray Instagram, 00:30:32.665 --> 00:30:41.161 which identifies the available hydrogennamic um, the available Ah drug in the in the whole test section. And you can see 00:30:41.161 --> 00:30:50.369 there are quite spike and well defined peaks, defining a a clear preference of the brown trout with both 00:30:50.369 --> 00:30:59.117 treatment, meaning that there seems to be no influence of Uh cylinder configuration, and also in the control case, 00:30:59.117 --> 00:31:08.785 preferences are ah quite high, especially for the small ah trouts and the wild trouts Which, were mixed ah 00:31:08.785 --> 00:31:18.070 in in the wild class they were small and large trouts. The fact that preference is higher for small crowds. It can be 00:31:18.070 --> 00:31:25.701 explained because more trouts need more to save energy, because they are smaller, so they have less powers when they are 00:31:25.701 --> 00:31:32.606 more interested in in saving energy than a very large trout that instead can swim almost undisturbed within the 00:31:32.606 --> 00:31:40.000 experiment, even in the control, a experiment. There is, we observe the same results. Interesting. The small trouts 00:31:40.000 --> 00:31:48.884 here. The blue curve actually has a peak towards even lower values of drug. And that is because, being smaller, they can 00:31:48.884 --> 00:31:56.130 go perhaps even closer to the lateral walls, where there are smaller velocities. And so um um smaller values of 00:31:56.130 --> 00:32:01.100 drag, something that a larger throat cannot do simply because of of the size. 00:32:02.369 --> 00:32:10.360 Now, this is the sheer stress, which was a big candidate. According to literature, there is the preference follows the 00:32:10.360 --> 00:32:18.406 estogram. So is not um very meaningful. In our opinion, @unoise@ and ah also the preference itself. The value ranges 00:32:18.406 --> 00:32:27.653 from one point two to maybe nine versus the Uh. They, they are much higher values of the drug function. And 00:32:27.653 --> 00:32:35.537 overall, there is not a consistency stint picture from treatment eight. It meant being in control, actually, for the 00:32:35.537 --> 00:32:43.873 sheer stress. A similar story for the turbulent kinetic energy which somewhat follows the patterns of the of the 00:32:43.873 --> 00:32:50.604 estograms. Uh, not much in the control experiments. But keep in mind that here the variations in Turban kinetic energy 00:32:50.604 --> 00:32:58.240 were not that big, as you can see that the range of Turban kinetic energy that is dictated by the cylinder experiments a 00:32:58.240 --> 00:33:06.837 lot larger than the control. So anyway, this seems to suggest a drug. Actually, the drug function explains a lot better 00:33:06.837 --> 00:33:15.799 than the other matrix space use. The question now is how drug compares to the Mb velocity. Okay, we may actually overdo 00:33:15.799 --> 00:33:24.558 that. Um again, on top we have the drug function with the preference curves on bottom the Ah. The um obtained from 00:33:24.558 --> 00:33:33.891 the from the Me velocity, which is the simplest matrix that is actually used in um in many models, um to be perfectly 00:33:33.891 --> 00:33:43.305 honest, also the mean works pretty well, although the results are a bit more noisy, eh? meaning that the drug 00:33:43.305 --> 00:33:53.067 function seems to perform slightly better, not significantly probably. And the explanation for that is that um um 00:33:53.067 --> 00:34:01.342 essentially the Sigma distant dwelling @unoise@ the the variants of the lateral and vertical velocity within the the 00:34:01.342 --> 00:34:10.451 test section were any way smaller than you square so. The difference between the drug function and and and 00:34:10.451 --> 00:34:20.255 you square or you, it is minimal. Yet there are some @unoise@ there. Are some um drawbacks associated with you that are 00:34:20.255 --> 00:34:29.017 that are overcome by the drug function, which seems to give @unoise@ um clear, cleaner results. Another questions that 00:34:29.017 --> 00:34:37.361 arises is why fish would choose this particular value of zero point twelve for drug @unoise@. Ah, we haven't 00:34:37.361 --> 00:34:45.313 investigated this much further. We haven't even ah normalized this value with something um. We think 00:34:45.313 --> 00:34:53.470 it may actually depend on the experiment. I mean, may be experiment dependent. So if you, if you managed to build 00:34:53.470 --> 00:34:59.560 another @unoise@ hyerodynamic ah field. Perhaps you would see that allows the trial to explore, even lower the drug 00:34:59.560 --> 00:35:07.797 areas, they probably might, but um, so I wouldn't. I wouldn't say that this is a a universal result, probably experiment 00:35:07.797 --> 00:35:17.601 dependent but clearly i Think these these results are enough to say that fish do like when low okay 00:35:17.601 --> 00:35:26.352 fish whenever they they are in the hydrogennamics influences fish behavior only on that. There is not much to talk 00:35:26.352 --> 00:35:36.217 about. So fish like to to save energy. Conclusions we say at the beginning that there is 00:35:36.217 --> 00:35:44.814 much debate on what is the most appropriate predict tore for space use @unoise@ ah because of the 00:35:44.814 --> 00:35:53.410 shortcomings we believe in in Ah, in the in the previous papers @unoise@ when ah, ah, specialized behaviors are ruled 00:35:53.410 --> 00:36:02.160 out, like in trainingment. Ah and so forth @unoise@ ah what we used the proposed drug function seems to explain 00:36:02.160 --> 00:36:10.420 why um seem Ah seems to be a robust, simple and biological relevant @unoise@ matrix to explain species @unoise@ 00:36:10.420 --> 00:36:19.919 and ah the other comment I would like to make is that, ah, similar drug function @unoise@ was used to explain not much 00:36:19.919 --> 00:36:29.831 space used by migratory ah roots a saw the use of these matrix can be extended to explain also the prefer roots by 00:36:29.831 --> 00:36:39.063 fish while migrating. An example is the work by of twenty twelve, where they have monitored a thing. The migration 00:36:39.063 --> 00:36:49.709 of of sturgeon upstream if you migrate upstream energy expenditure is very, very relevant problem. Um, think 00:36:49.709 --> 00:36:57.574 about a fish that has to do thousands of kilometers of hundreds, depending on the species, any delay, any over energy 00:36:57.574 --> 00:37:05.688 spent will would delay the reaching the spawning habitats, or actually would prevent the fish to to reach them. So it is 00:37:05.688 --> 00:37:14.140 very important to save energy. And they have observed that the sturgeons like to migrate on the side of the banks where 00:37:14.140 --> 00:37:22.469 velocities are very low, and presumably drug is very low, and therefore energy um expenditure. And this ah makes a lot 00:37:22.469 --> 00:37:30.543 of sense as we have discussed. But I will leave you with an open question. So what about downstream migration? whenever 00:37:30.543 --> 00:37:37.339 you have downstream migration, you could actually be drifted nicely by the water. So energy expenditures seem 00:37:37.339 --> 00:37:45.889 intuitively should be less @unoise@ less important @unoise@ and this is something that we will address for years 00:37:45.889 --> 00:37:55.539 after the Eh, the break um of ten minutes. Okay, thank you. Any question, anything? 00:37:55.659 --> 00:38:15.813 okay, let us get back to fish, not grilled fish. But we are now on the second part of this. Uh of the slot. And we have 00:38:15.813 --> 00:38:22.680 finished the previews @unoise@ talk about um, uh, talking, about uh, what about downstream migration? what are the 00:38:22.680 --> 00:38:30.861 preferential roots? is there any? well, we said the the drugs shouldn't play any role, because now fish are going 00:38:30.861 --> 00:38:40.676 are drifted downstream, essentially. So definitely, Energetics should be less less important. We carried out a 00:38:40.676 --> 00:38:50.108 study about eels downstream migration clearly, all this studies. Ah, whatever you find, if species dependent, we 00:38:50.108 --> 00:38:58.152 chose to study eels because it is an endangered species. And Ah, that the, the, the ill stocks around the world have 00:38:58.152 --> 00:39:05.379 have depleted heavily in the past thirty years, and it is considered now an endangered species. Officially, all the work 00:39:05.379 --> 00:39:13.567 that I'm going to present was most of it actually not not all of it, but most of it is containing this paper, which is 00:39:13.567 --> 00:39:20.259 published in the Proceedings of Our society. B and I want to acknowledge the contribution of Adam Piper Post Dock at 00:39:20.259 --> 00:39:28.014 that time with the Postdock in Southampton. Fabulous and and a professor at the University of Padua, 00:39:28.014 --> 00:39:36.549 Rosalie Rose, right from the U. K and, Again Paul Camp and my colleague, a former colleague in Southampt. 00:39:36.869 --> 00:39:46.983 Um, let us talk a little bit about eels. Ills are mystery animals. Ah, the the life cycle is is Ah, is really incredible 00:39:46.983 --> 00:39:55.383 to me. When I read about about is I was really, really surprised. Ah, do you know about that? have you ever heard?"" 00:39:55.383 --> 00:40:05.580 Yeah, somebody says," Yes, one, okay, so let me just summarize the life cycle of of ills. All the years in Northern 00:40:05.580 --> 00:40:15.453 Europe and North America are start their life in the sargassum, essentially, and um @unoise@ as far as from the eggs 00:40:15.453 --> 00:40:26.266 they become very good thank you < laugh > And then they, they, they are born here, 00:40:26.266 --> 00:40:35.669 and then they start being drifting by currents, essentially, and going back to rivers and lakes in North, while in 00:40:35.669 --> 00:40:45.394 Europe and in North America. By the time they arrive in fresh waters. They are glass eels. That is when they start um ah 00:40:45.394 --> 00:40:53.147 migrating upstream while they migrate, and they live in the in the fresh water, they become Elvis and then yellow eels 00:40:53.147 --> 00:41:01.270 so they they can grow larger and larger, until they they get this silver color become what is called silver real . 00:41:01.280 --> 00:41:09.699 And then they have instinctively the need to to migrate downstream and to go back to the ocean. And what they do is 00:41:09.699 --> 00:41:17.899 they, they, they travel apparently all the way again to the Sargasso Sea to for for formating, essentially, which is 00:41:17.899 --> 00:41:26.105 this desponding with the supposed spawning area. But nobody has ever proved it. As ever tracked eels from North America 00:41:26.105 --> 00:41:34.193 or Northern Europe to the Sargasso Sea. There is there. There is a very recent paper, I think. Yeah, two thousand and 00:41:34.193 --> 00:41:42.084 sixteen. A big group of scientists who managed to track the fields that were released in Ireland, in Uh, in 00:41:42.084 --> 00:41:50.556 Sweden, in Norway and in France and Spain as well, and they managed to track them with some fancy techniques. And the 00:41:50.556 --> 00:42:00.424 trajectories seems to suggest that all days are actually going to the sagacity. But um, you know, they lost signal of 00:42:00.424 --> 00:42:10.728 all these fish while while moving. Um @unoise@ what we are interested in in this talk are so what we are 00:42:10.728 --> 00:42:20.031 going to talk about are big ills roughly one meter long that are moving downstream in fresh waters. That is what 00:42:20.031 --> 00:42:26.179 we are. Um aiming eh to eh what I'm aiming to to show you today. 00:42:26.719 --> 00:42:36.059 Um, as I say, there is the severe decline of the European ill, because there are um while migrating both upstream when 00:42:36.059 --> 00:42:44.549 they are glasses or downstream when they are silver, is the migration is delayed or blocked by by river infrastructures. 00:42:44.550 --> 00:42:55.906 I was showing, like, dams, weirs and so forth ill suffer really high rates of mortality, because they are susceptible to 00:42:55.906 --> 00:43:03.359 impingment at exclusion screens. Whenever you have a water intake or something. The has to put some sort of 00:43:03.359 --> 00:43:10.809 screens to prevent the passage or to prevent the passage of debris and in fish itself, and they get in pinch. They 00:43:10.809 --> 00:43:17.605 don't have the power. If the frost is high, they don't have the strength to swim back. And and so they get stuck in 00:43:17.605 --> 00:43:18.169 these screens. 00:43:19.340 --> 00:43:26.680 Um, there is general lack of knowledge about his behavior in correspondence of manmaid structures like dams and weirs. 00:43:26.690 --> 00:43:35.135 But it seems that that ills are what is called so that they like to go there to touch the structure, possess what 00:43:35.135 --> 00:43:42.807 is going on, and then take a decision based on that I saw. They were more prone to impangement, because if they didn't 00:43:42.807 --> 00:43:51.307 and got close to the screen. Probably they wouldn't be impinged generally, all the downstream passage solutions that are 00:43:51.307 --> 00:44:00.497 tried worldwide have variable efficiency effectiveness, but in generally slow this is an emergency we need 00:44:00.497 --> 00:44:09.688 to understand how fish our ills actually behaving correspondence of this structure so that we can design these 00:44:09.688 --> 00:44:20.734 structures in a way that is friendly for for ills again, with the same concept as explained this morning. We need 00:44:20.734 --> 00:44:30.333 to understand what really attracts ills so that we can guide them towards a safe passage. And what repels is that we can 00:44:30.333 --> 00:44:39.256 find ways to repel them from dangerous areas like watering takes for turbines or or or screens @unoise@ Um. This is a 00:44:39.256 --> 00:44:47.308 very sad picture taken from the literature of an example of impingment in a in a trash track of many hills, and you see 00:44:47.308 --> 00:44:56.052 how many there are. And besides, in a water intake. And besides being very, very sad, this is also a tremendous and high 00:44:56.052 --> 00:45:06.121 cost for the, for the for the water companies, because these have to be. These trucks have to be cleaned daily, and the 00:45:06.121 --> 00:45:15.585 screens get get clogged. And so in dams, they lose head that converting energy and so forth. So beside being necological 00:45:15.585 --> 00:45:23.333 emergency results, an interest by by industry to get rid of these ills because they cost money for them. Okay, so it is 00:45:23.333 --> 00:45:31.222 a very, very important problem we want to solve. We want to save the use to do that, we need to get an understanding on 00:45:31.222 --> 00:45:32.520 of how they behave. 00:45:33.090 --> 00:45:42.612 Um in an attempt to do that and to understand what they do. I already telling you, I will give no answers today. 00:45:42.612 --> 00:45:51.271 Essentially, the outcome of this study is that we don't know, but we have some new questions, which which hopefully will 00:45:51.271 --> 00:45:59.433 are sharpened with respect toward the previous literature. I was saying in an attempt to understand how our use 00:45:59.433 --> 00:46:10.094 react to a hydrogennamic Steamboat. We decided to carry out some field experiments in a very, very special place. This 00:46:10.094 --> 00:46:20.820 is a channel located in southern England The. Channel is mm. Okay, the flow direction here is from left to 00:46:20.820 --> 00:46:29.240 right. Ah, there is um @unoise@ Ah. Previously, this was Ah. This was a water intake for for hydropower generation now 00:46:29.240 --> 00:46:31.240 is mayor was made redundant. 00:46:31.389 --> 00:46:41.270 But still there is a barrack in here which, is sloped which, can be used to construct or open the flow essential 00:46:41.270 --> 00:46:49.610 you can. You can alter the hydrogennamics as you want using this. This did this barrack. 00:46:49.769 --> 00:46:58.520 And this is the bathroom tree. The color scale is in here. It is in water. Depth would have ranged from one point eight 00:46:58.520 --> 00:47:07.417 meters to zero point fifty. Four, the flow. There is nothing in here. The floor is all diverted in there. We did. We 00:47:07.417 --> 00:47:17.018 decided to essentially do two experiments. One, when the the gates of the barrack were fully open, which implied very 00:47:17.018 --> 00:47:18.039 low accelerations. 00:47:18.349 --> 00:47:26.080 And the other experiment we call it constricted, where we have constricted the floor to create, to create very strong 00:47:26.080 --> 00:47:30.399 acceleration in proximity. Excuse me of day of the barrack. 00:47:30.599 --> 00:47:34.340 What we did essentially was to put some ills. 00:47:35.880 --> 00:47:44.720 And here, farther upstream from the rake and monitor, uh, their movements are show you how, uh, until they escaped. 00:47:44.730 --> 00:47:55.558 Okay? and we monitor that, and they fully open, which is you. Well, ah, treatment and constricted Ah conditions. That 00:47:55.558 --> 00:47:58.610 is, that is what we did. 00:47:58.760 --> 00:48:10.200 The nice thing is that we could carry out the experiments by keeping the same, the same discharge. 00:48:11.230 --> 00:48:17.715 Uh, so the volume of water, uh per unit second. Uh flowing in the river, because there were upstream here. There was a 00:48:17.715 --> 00:48:23.916 wheel that allowed to divert some water when some natural of the river would put the discharge higher so we 00:48:23.916 --> 00:48:28.990 could actually control it very, very nicely. It is an excellent field site, because we can control the hydraulics. 00:48:30.750 --> 00:48:39.631 I was saying before, uh, doing the use experiments. We did some transact with um, an instrument called that 00:48:39.631 --> 00:48:49.903 allowed us to measure mean and few cross sections of the channel. And these are the the black line. And just a 00:48:49.903 --> 00:48:59.456 little bit of background and an is a big instrument with a huge sampling volume. Okay, so what that actually 00:48:59.456 --> 00:49:08.776 measures at when flow depth are this shallow is is the bike is the mean velocity. It is the depth average. They were 00:49:08.776 --> 00:49:17.223 lost. Nothing else or nothing, really, uh, nothing really fancy, no turbulence whatsoever. Okay, but we could at least 00:49:17.223 --> 00:49:27.724 map, uh, the flow. Uh, fairly well. Um, we could. We also put some maintenance in here in these Red Cross sections, um 00:49:27.724 --> 00:49:35.683 to monitor, essentially at which that eels were swimming. Usually they swim very close to the bottom. We wanted to 00:49:35.683 --> 00:49:37.810 verify that and actually adopt that. 00:49:37.969 --> 00:49:50.838 That was the case um. We used twenty use per treatment. So twenty years for the fully open experiment and full and 00:49:50.838 --> 00:49:59.376 twenty years for the constricted experiments. I forgot @unoise@ ah to mention that we want to create a sort of an area 00:49:59.376 --> 00:50:05.991 with high and low accelerations, because according to literature, acceleration @unoise@ is is um I'm talking about 00:50:05.991 --> 00:50:13.396 attackive acceleration. Okay, so the flow is steady, so it is the acceleration due to the @unoise@ today to the % uh 00:50:13.396 --> 00:50:16.790 of the flow. Okay, that is what I'm talking about. 00:50:17.250 --> 00:50:26.442 I was saying, it seems that this acceleration is what other fish species respond to, for example, the the work of 00:50:26.442 --> 00:50:34.835 Goodwin and and coworkers show that, uh, really um, salmon. It is respond um to acceleration. Uh. I mean, the the 00:50:34.835 --> 00:50:42.844 response is pretty clear when the acceleration is low, they tend to mill when it is very high there. They, they, they, 00:50:42.844 --> 00:50:48.230 they get scared. They make some rejections. They go back upstream, actually, or downstream, depending where they are 00:50:48.230 --> 00:50:48.530 migrating. 00:50:48.769 --> 00:50:56.389 So our hypothesis was, is acceleration the right metric also that determines some some behaviors to wheels as well. 00:50:56.400 --> 00:51:04.870 And the idea was to test these hypothesis by creating low and high acceleration zones here to see the response of 00:51:04.870 --> 00:51:11.840 of eels. I was saying with it, we used a total of twenty years per treatment experiment again, were carried out at 00:51:11.840 --> 00:51:18.353 night. Not much needed, because visibility in water was very low anyway, but we decided to do all the experiments at 00:51:18.353 --> 00:51:21.449 night, which is also when eels are very, very active. 00:51:22.090 --> 00:51:33.049 We use acoustic tugging to monitor eels to tack ills. When when swimming in the in the channel, essentially what you do. 00:51:33.059 --> 00:51:42.793 You put a lot of microphones in here. You put a tag under the skin of the hills, and the microphones are able to to 00:51:42.793 --> 00:51:50.035 detect the signal. And by triangulation, I think they work out. You can work out the trajectory the accuracy is 00:51:50.035 --> 00:51:57.913 @unoise@ after a little bit of an Isis we worked out. It was roughly half a meter between half an hour. Definitely less 00:51:57.913 --> 00:52:07.002 than a meter. Okay, not great, but not. But keep in mind that just to give you a scale. These are ten meters. So the 00:52:07.002 --> 00:52:17.263 channel is fifteen twenty meters wide, and they had to to swim through length of what, and were thirty, forty meters in 00:52:17.263 --> 00:52:27.899 total. Ok um we verified with this sorry. That is swarm at the bottom, which is an important Ah, which is 00:52:27.899 --> 00:52:36.262 important information for what I'm going to tell you a very soon to back up the measurements. The velocity 00:52:36.262 --> 00:52:44.639 measurements, we decided to do some very simple basic and numerical simulations of the flow within the channel @unoise@ 00:52:44.639 --> 00:52:54.339 so that we could uh have a much more detailed flow description of the field site @unoise@ we use the velocity 00:52:54.339 --> 00:53:02.725 measurements that we did to calibrate these model Uh models We. Did two dissimulations Uh, uh. So essentially, 00:53:02.725 --> 00:53:13.608 what we are simulating our uh flow depths with a very rough K. Absilon closure scheme, um. And these are the results we 00:53:13.608 --> 00:53:22.520 obtain. This is the velocity field that the in here. And this is the acceleration field, which I have already told 00:53:22.520 --> 00:53:32.788 you seems to be an important matrix to interpret eel's behavior um as you can see, the the fully open treatment has a 00:53:32.788 --> 00:53:42.163 fairly smooth exit from the channel, whereas the constricted um treatment has a higher um um sorry zone a very, very 00:53:42.163 --> 00:53:50.607 high. The lost in proximity of the barrack @unoise@ and. Ah of, course a, high ah compactive acceleration, um they I 00:53:50.607 --> 00:53:58.879 forgot to put, uh, a picture, uh. I mean, the the agreement between the measurements and the Uh and the dissimulation 00:53:58.879 --> 00:54:08.136 was not bad at all. We managed to capture fairly well the flu depth and the mean velocities, except for a, you know, for 00:54:08.136 --> 00:54:15.040 a few areas here, we had some problems and a little bit here, but overall, very good agreement. 00:54:16.570 --> 00:54:22.320 I always say you can do very good simulations if you have very good measurements so that you can tune your more than as 00:54:22.320 --> 00:54:24.960 you want. There is an extension but that was the case. 00:54:25.230 --> 00:54:30.639 And we use the model only to refine the the flow description. 00:54:32.619 --> 00:54:42.602 Now, to interpret um the behavior of ills. We needed to define some sort of of behaviors, essentially how what is 00:54:42.602 --> 00:54:46.019 behavior? you need to to quantify it somehow. 00:54:46.239 --> 00:54:55.140 And first we looked at at the the ill trajectory. And as I will show you what they did was basically swimming almost 00:54:55.140 --> 00:55:02.757 drifting passively in this first part and then once they reached they started doing some crazy things. We 00:55:02.757 --> 00:55:10.230 started going back start milling @unoise@ so. We wanted to interpret @unoise@ mm the behavior in this area where things 00:55:10.230 --> 00:55:18.797 got a little bit less smooth for them @unoise@ % um so we defined % uh switch points, switching behavior. Okay, from A, 00:55:18.797 --> 00:55:27.536 from A, A, a passive drift behavior to something that is much more active with a fine rejection. When dance removing 00:55:27.536 --> 00:55:34.340 fish changed, switched abruptly from negative to positive. Meaning following the current going back essentially. 00:55:34.349 --> 00:55:40.710 So it changed direction completely. That is what that means. And moved in a direction for at least three meters. 00:55:41.030 --> 00:55:48.861 You may say this is a subjective definitions. I agree. But what do you do? it is very difficult. And this is a 00:55:48.861 --> 00:55:56.330 definition that is used to describe fish behavior in the literature. So that was our best shot, at least consistent with 00:55:56.330 --> 00:55:57.040 previous studies. 00:55:57.940 --> 00:56:05.910 We also defined exploratory behavior when downstream moving from switched from negative to real taxes to some sort of 00:56:05.910 --> 00:56:13.461 lateral movements greater than three meters length, perpendicular to the stream wise flow, and at least, and compassing 00:56:13.461 --> 00:56:24.229 more than two tarts. You know, it is < laugh > It is very arbitrary, but ah you know, you can tell by I, you can 00:56:24.229 --> 00:56:32.958 interpret the behavior of of a fish just by looking at the trajectors, but quantify switch points in a rigorous way. Non 00:56:32.958 --> 00:56:43.250 arbitrary it is very difficult. So this is the best we could do by matching intuition with with quantification 00:56:43.250 --> 00:56:52.297 @unoise@ sorry yeah it is definitely subjective. It is probably, if you give the same data to somebody else, 00:56:52.297 --> 00:57:01.781 they would see something else. But this is. This is one of the problems But. Anyway in, my opinion, one of the, if 00:57:01.781 --> 00:57:12.623 not the best result of these experiments was not much about the switch points was how he is swam upstream and look 00:57:12.623 --> 00:57:21.108 at these trajectories Thirty seven over forty years actually moved downstream trees decided to do whatever they 00:57:21.108 --> 00:57:32.277 wanted. They went upstream. We lost them, but thirty seven. So a good percentage swam went out through the barracks, and 00:57:32.277 --> 00:57:43.943 all of them, all of them swarm in these two corridors in the channel with a velocity. There was about zero point seven, 00:57:43.943 --> 00:57:52.864 zero point six, the water, the depth average, the lost now except for this ill here that actually patched the side 00:57:52.864 --> 00:58:03.550 banks. All of them did not came into into contact with the banks, and yet they decided to go into these two corridors. 00:58:03.559 --> 00:58:13.456 None of them went through the centre. Why weird < laugh > Very, very strange. Um, I was saying the velocity was zero 00:58:13.456 --> 00:58:21.230 point six with a point seven, the the depth of his velocity, which makes sense, because they swarm at the bottom, where 00:58:21.230 --> 00:58:29.737 blossoms are slightly lower, and they were drifting by the the near wall below state. So this is the first interesting 00:58:29.737 --> 00:58:40.621 results that we will discuss later on now, thirty, five out of thirty, seven years. So all of them showed a reaction 00:58:40.621 --> 00:58:51.425 when they approached the rack. Okay, now you can tell your definitions of of rejection, of of for exploratory. How did I 00:58:51.425 --> 00:58:58.224 take call it exploratory behavior is is arbitrary. You can see whatever you want. Fine, but it is definitely an 00:58:58.224 --> 00:59:05.567 objective result. All of them reacted. All of them stopped being drifted. So said I, they either stopped or or got 00:59:05.567 --> 00:59:18.927 scared. Okay, interestingly in the open, a fully open experiment. Seventy five percent of the yields showed 00:59:18.927 --> 00:59:19.930 exploratory behavior. 00:59:20.460 --> 00:59:22.250 Okay, not rejection. 00:59:23.789 --> 00:59:33.712 Whereas in the constricted behavior, all these went nuts. So one day, they approached the bar like they started swimming 00:59:33.712 --> 00:59:42.836 upstream and doing really, they were clearly scared um every time there was a rejection. Nearly ninety one percent 00:59:42.836 --> 00:59:48.550 of all the fish they rejected. Did that multiple time. So they tried to pass. They got scared. They went back. 00:59:48.559 --> 00:59:53.755 They swarm a little bit. Then they went back there. They got scared again. Then there was some habituation, and 00:59:53.755 --> 01:00:01.792 eventually they went through the rack. Okay, so that that was what happened just to give an idea of what exploratory 01:00:01.792 --> 01:00:10.695 behavior and rejection looks like. So in the constricted case, the ill moved here, then turned, decide to turn, and then 01:00:10.695 --> 01:00:19.293 felt something, and got scarce. Ah, when they went upstream really far away. Look at Ali. Went almost back to to where 01:00:19.293 --> 01:00:27.377 it was really released, and then eventually tried again and then, and then escaped um in the constricted . Well, 01:00:27.377 --> 01:00:34.739 let me describe the the fully open experiment showed ills drifting away, drifting away. And then when they came here, 01:00:34.739 --> 01:00:43.422 most of them started doing. This is sort of milding behaviour. Well, should I go true? should I not seems, seems a bit 01:00:43.422 --> 01:00:51.953 dangerous. I don't know." and then decided to go okay, we did some sort of analysis of these trajectories and results 01:00:51.953 --> 01:00:59.605 were pretty intuitive. There were longer posts which distance traveled in the constricted case and much longer 01:00:59.605 --> 01:01:08.408 than the than the fully open. And there were much higher post switch velocities of the hills. So in the constricted 01:01:08.408 --> 01:01:16.508 case, they swam away really, really fast, indicating that they got scared for some reason. Okay, they felt the 01:01:16.508 --> 01:01:19.920 constricted area as a dangerous era. 01:01:20.309 --> 01:01:31.191 However, this kind of makes sense. This is consistent with the picture. Accelerations scare well eels respond to 01:01:31.191 --> 01:01:45.389 accelerations. However, um it is not clear which whether they responded really to acceleration or something else. 01:01:45.750 --> 01:01:55.338 If you mapped all the switch points for the fully open and the constricted case. And they all occurred in proximity of 01:01:55.338 --> 01:02:04.159 the intake. It is a bit more spread. The case were the constricted experiment. But again, even if you allow for um, some 01:02:04.159 --> 01:02:12.900 sort of flexibility on how you define the behavioral. A switch points none of them really was in the areas of high 01:02:12.900 --> 01:02:16.480 acceleration, all the switch points were kind of upstream. 01:02:17.929 --> 01:02:26.740 So the eels clearly did not feel these accelerations of if they, if they managed to detected the acceleration field. 01:02:26.750 --> 01:02:32.030 They use the system. I don't know. I have no idea about. 01:02:33.079 --> 01:02:40.797 So this opens the questions. Did they respond acceleration like salmon, it is too, or to something else if they 01:02:40.797 --> 01:02:47.787 responded to acceleration, how? how did they detect the acceleration itself? because they did not came into contact with 01:02:47.787 --> 01:02:48.149 it. 01:02:49.340 --> 01:02:58.241 So very, very weird. Perhaps they responded to noise, you know, clearly is a much more noise environment. When you 01:02:58.241 --> 01:03:05.581 construct the floor. But Adam did some experiments with by by mimicking the noise of a waterfall. The noise you would 01:03:05.581 --> 01:03:14.500 get in the constricted case with some speakers and microphones, you managed to tune it and doesn't seem to be the the 01:03:14.500 --> 01:03:21.973 driving matrix @unoise@ pressure gradient could be another another another option, or we. Perhaps it was a short time of 01:03:21.973 --> 01:03:33.012 the experiments. If you go back % um. If you go back to the to the to the map in here. There is sort of a scoured 01:03:33.012 --> 01:03:40.348 region in here, which is with respect to background, probably thirty, forty centimeters. There is like a slight um 01:03:40.348 --> 01:03:48.549 deepening of the of the of the bottom. Perhaps that generated some sort of flow separation, or something that would 01:03:48.549 --> 01:03:56.885 stabilize use I don't know. It seems unlikely, because it really occurred over large longitude in our scales. So 01:03:56.885 --> 01:04:03.949 it is not my preferred explanations, not that I have one, but I think this is. This is unlikely it is a possibility. 01:04:05.880 --> 01:04:13.396 Um, so yeah, Gabby was suggesting pressure gradients. But again, we from the um ah, from the modeling and the 01:04:13.396 --> 01:04:20.987 measurements of the of the flow depths, the, the, the, the gradient in day in the water surface profile was not that 01:04:20.987 --> 01:04:28.924 severe in corresponding to the in correspondence of the switch points, uh, or if it was well, we don't have a way 01:04:28.924 --> 01:04:36.609 to quantify it. So < laugh > I can't really tell, but it didn't seem definitely there was not ah like, 01:04:36.609 --> 01:04:44.256 ah, ah, a sharp change. Okay, pressure. Great. And so we checked that. But mm, I'm not sure is the right answer @unoise@ 01:04:44.256 --> 01:04:52.202 Oh, another colleague of mine is convinced it could be pressure to hide in. So we need to dig farther on that @unoise@ 01:04:52.202 --> 01:05:00.090 um. The other interesting result, which Ah, which is my favorite in this study is that, Ah, as I've shown all the ills 01:05:00.090 --> 01:05:07.269 liked to swim @unoise@ over these two corridors in the upstream. Let us call it in the undisturbed part of the channels. 01:05:07.960 --> 01:05:13.249 Um none of them decided to to swim at the, you know, at the in the center. 01:05:14.329 --> 01:05:24.269 The question is, are two questions. Why did they choose to swim duck close to the banks. And if they, I mean, they they 01:05:24.269 --> 01:05:27.169 were swimming at night. Conditions zero visibility. 01:05:27.639 --> 01:05:30.919 They did not come into contact with the latteral banks. 01:05:31.809 --> 01:05:42.500 If they decided so they, they wanted themselves to swim, close the banks, how on earth they detect the banks? 01:05:43.309 --> 01:05:50.040 how did they know that there were banks over there, because it was completely night time, and they did not touch banks. 01:05:50.050 --> 01:05:57.768 So the current hypothesis that we have is that the proximity to latter, our boundaries could be a way as lie, like an 01:05:57.768 --> 01:06:05.487 avigational cue. So they know what the boundaries are. They know they are swimming in a channel in a river, and they are 01:06:05.487 --> 01:06:13.605 following the main current. That is what that is our explanation. But yet again, the second question is, eels then 01:06:13.605 --> 01:06:23.872 probably use the system. The latter aligned to detect some sort of hydrogennamic footprint of these, or of some 01:06:23.872 --> 01:06:31.800 sort of a signal signature they could detect to see, to feel the lateral banks. 01:06:32.429 --> 01:06:40.410 And to me, the only well what generates a signal that comes from from the latter balance towards the center, towards 01:06:40.410 --> 01:06:47.469 where they were swimming are perhaps the secondary current. So perhaps they are able to sense secondary currents, or the 01:06:47.469 --> 01:06:54.691 sheer stress that they generate at the bottom. Remember that they like to swim close to the bottom. So perhaps they are 01:06:54.691 --> 01:07:02.430 able to to sense some announced friction um with respect to the, to the, to the Channel Center. I, I really don't know. 01:07:02.440 --> 01:07:10.421 There was um. There was an hypothesis. We threw in the table when we were right in the paper. I don't think it is 01:07:10.421 --> 01:07:11.629 written in the paper. 01:07:13.559 --> 01:07:24.429 So just to conclude, um is seems to respond to either dynamics. There is no doubt on that. But the the dynamic matrix 01:07:24.429 --> 01:07:33.934 that respond to is not clear. Actually acceleration alone. As for salmon, it is, uh, doesn't seem to be a candidate. He 01:07:33.934 --> 01:07:38.149 will swim close to lateral banks. But how do they detect them? 01:07:38.380 --> 01:07:46.042 I don't know what is the der dynamic senior. They they manage to identify I don't know. And these are 01:07:46.042 --> 01:07:53.299 really, these are open questions and in my opinion, are really, really important imagine that we understand. 01:07:54.000 --> 01:08:00.500 So with these results, we know that they like to swim close to the to the channel banks. 01:08:00.679 --> 01:08:06.691 So this is something that attracts the ears, remembered at the beginning. We say we need to understand what attracts our 01:08:06.691 --> 01:08:12.339 repels. So here we have the two ingredients. We know that the constricted case repelled more than their fully open case. 01:08:12.349 --> 01:08:19.445 So we have some data showing that, and we know that they really like to swim close to the banks. So if they really like 01:08:19.445 --> 01:08:25.824 to swim close to the monks. What is the hydrogennamics that attracts them. If I manage to, if I managed to reproduce 01:08:25.824 --> 01:08:32.177 that artificially by putting like stripes of roughness. For example, in proximity of the damp. Perhaps I can fool the 01:08:32.177 --> 01:08:36.970 yields and guide them towards a safe passage creating artificial secondary currents. 01:08:37.569 --> 01:08:46.010 And if I understand what what we really repelled ills from Ah in the barrack? if so, I really know, what is the 01:08:46.010 --> 01:08:53.261 matrix that that Ah, that that generated these rejections, then I can ah take. I can find hydrogennamic ways to 01:08:53.261 --> 01:09:02.162 implement them in proximity of dams, and therefore, ah, prevent used to go to screens or to be trimmed into the into the 01:09:02.162 --> 01:09:10.074 term. So I think, ah, this old study. Ah, really integrate fundamental signs with clear engineering applications. Um, um 01:09:10.074 --> 01:09:18.920 and ah. So to me is really, really interesting. And besides the the paper. Ah, that I have already. Ah, mentioned 01:09:18.920 --> 01:09:28.214 in our paper. There are also a couple of more references to look at that. I have used to for this presentation. And for 01:09:28.214 --> 01:09:37.119 me, this is the final slide. I thank you very much for your attention. And ah, as usual, if you have questions, filthy.