| (Click on a 'Start Time' to view the video) |
| 00:02 | yeah I think it seems like many have joined so so let's go ahead and and get started so I would like to welcome everyone to this 18th webinar in this ongoing discussion series on tipping elements irreversibility and upoad changes in the air system my name is AAS Muhammad I'm a post doctor of research scientist at the University of Hamburg part of the functional Forest ecology group um also supporting Earth commission working group to on biosphere interactions and I'm glad to be your moderator uh today and just a quick |
| 00:44 | reminder uh specifically for those who are joining us for the first time so the goal of this discussion series is to advance knowledge on tipping elements support efforts to increase consistency in treatment of tipping elements in the scientific community help develop a research agenda and design joint experiments and ideas for a tipping element model interc comparation project or tip map for short and there are a huge like network uh behind this series which are uh the Earth commission so Earth commission is |
| 01:21 | a global team of scientists which with a mission of to to Define safe and just Corridor for people and uh planet we have a names uh the analyzes integration and modeling of the earth System project which is an international network of Earth system scientist and Scholars that seek to develop Innovative interdisciplinary ways to understand the complexity of the natural world and its interactions with human activities so both ases and uh Earth commission are hosted by Future Earth which is itself a Global Network of scientists researchers |
| 01:58 | and innovators collabor ating for a more sustainable planet and the third network behind this series is the safe landing climate Lighthouse activities of the world climate research program or wcrp for short so which act also to explore uh roots to safe landing spaces for human and uh natural systems and this event is organized or co-organized by The International Institute for Applied uh system analyzis yasa The poam Institute for climate impact research and the global system Institute at University of exitor so our |
| 02:39 | Focus today is on ecological tipping points and resilience when it may occur and when not and uh briefly our agenda for today is that we will have two presentations uh each about around 20 25 minutes long followed by a discussion session which will be also around 30 minutes and both of our speakers will present before the Q&A session so please during their po their talks feel free to pose questions and indicate to whom you are directing this question uh using the Q&A feature at the bottom of the zoom |
| 03:19 | window and uh there is also the option that you vote for your preferred questions so don't hesitate to do that and we will then get to this question question in the discussion session following immediately the presentations and one last point is that following the event today so we will post the recording and the slides on uh the Tipping Point series confetti web page for future accesses so we have an exciting time ahead and we are looking forward to also your interaction in the Q&A box so I will start with our uh |
| 03:59 | first speaker today and so our first speaker is Juan roacha Juan is a research scientist at the Stockholm resilience Center where he Coles the team on complex ecosystem his research questions are oriented to understanding critical transitions from regime shifts in ecological systems to Collective action in in society he's currently focusing on the ideas of cascading effects how for example example a critical transition in an ecosystem in the world can impact the likelihood of other ecosystem tipping over so Juan |
| 04:39 | will share insights on how ecosystems are showing symptoms of resilience loss so we are thrilled to have you uh with us today Juan so the floor is yours thank you vas uh thanks also to the organizers for inviting me to present some of my work uh please confirm if you see my screen on full full view of the slides yes super thanks a lot uh so the the talk I'm presenting today is based on a paper that was publishing Environmental Research letters in 2022 um the slides are on my website and just if we go quickly to the |
| 05:15 | last slide uh you will have find there the the link to the slides the link to the papers and everything on the on the presentation that is blue are active links uh if you're interested in Reading More on on some of the methods or data sets that I'll be using uh so the title of the paper is ecosystems are showing symptoms of resilience loss uh being thinking of resilience in the context of regime chips which broadly speaking they are defined as large abrupt and persistent changes in the function and structure of ecosystems canonical |
| 05:43 | examples include the CH from forest to Savana this picture was taken uh in the chako region of Latin America where parts of the U Forest is being transformed to Savannah due to cattle raning activities another canonical example is the chi from Coral dominated reefs to macro Al of soft coral reefs has been documented in uh many places around the world with impacts on communities that depend on these ecosystems in terms of Tourism uh maintaining fish communities or Coastal erosion another canonical example is the |
| 06:14 | collapse of Fisheries it can compromise the employment of millions of people who depends on that type of livelihood but also our capacity to produce food so as you see this events or this type of phenomena occuring very different spatial and temporal scales but what do they have in common is that they are dominated by nonlinear Dynamics and tipping points um when those tipping points are are reach is often the case that the ecosystem structure change in configuration and then the services that we receive from that ecosystem change |
| 06:45 | sometimes um in in a in a negative manner uh because they are nonlinear Dynamics they are very difficult to predict uh sometimes hard or impossible to reverse and um um for those reasons are very important to better understand when in the world they likely to occur but answering that question depends on our ability to observe and measure resilience uh a number of groups around the world have been trying to focus on that question and using a different data sets and and techniques uh in this paper I try to combine some of those uh recent |
| 07:17 | developments on Earth observations so the data I use uh was uh harmonized by the Earth system data lab uh they are based at The maxman Institute of biog geochemistry but what they're doing is taking remote sens in prod from the European Space Agency NASA NASA and the like and putting them on the same space and temporal resolution uh they have about 75 variables mostly uh reflecting changes in climate as so so there are about two or three decades of weekly observations on how the planet is doing but some of those variables have to do |
| 07:49 | with how ecosystems are doing in particular I was focusing on Gross primary productivity consistent respiration and Le area index has three data sets that speaks of how the terrestrial biosphere is changing and also the color of the oceans or or which is appr proxy of chlorophile a so primary productivity on the SE surface um put it together they are about 1 million pixels and a resolution of a quarter degree uh grip resolution and weekly observations uh depending of the data set we have between 18 years or two |
| 08:21 | over two decades and with this time series then we can start playing around with some of the uh techniques that people have been using to approximate resilience the first one or the first family of um metrics is critically slowing down uh it's a method that expect uh resilience loss due to the um flatness of the Basin of Attraction so resilience in this mathematical framework is the size of the Basin of Attraction you can lose size when the Basin of Attraction get shallower critical slowing down is a particularly |
| 08:51 | good method where the you are approaching the tping point by increasing slowly on some controll variable or drive um then what you expect from time series or the signal that it lives is increasing barriers and autocorrelation dakos here in this paper in plus one describes very well the methods for um time series and Sonia key for special early warnings U however critical slowing down doesn't perform very well when the Dynamics are driven by stochastic processes or noise um and more recently uh Titus and Watson |
| 09:20 | proposed critical speeding up which is uh the case when the Basin of Attraction is not getting shallower but it's getting narrower in that case the system becomes more vulnerable to noise or stochastic variabilities and what you expect statistically is an inre a decrease sorry on variance and autocorrelation as a complementary metric I also use the fractal Dimension uh that's a metric that is commonly used in engineering and Medicine uh to detect vulnerability of systems to collapse uh is a measure of similarity across scales |
| 09:49 | but typically is interpreted as appr proxy of adaptive capacity so your doctor when you go to an electrocardiogram or or checking your heart they check the fractal Dimension and if the that fractal Dimension is high that means that your body can deal with a number of different stresses and you will perform well if you are running or going upstairs hiking up in the mountains but your fractal Dimension is low that means that your heart might have a struggles at dealing with the different situations where you can put |
| 10:15 | your heart and then you are vulnerable to think like a heart attack or something like that so that's how it's interpret the diversity of um behaviors that a system has to deal with the different types of disturbances and the bottom you see some of the um Theory and statistical methods to measure the fractal Dimension right uh if we go to one pixel um here we have on the left uh the case of one particular pixel for data on Gross primary productivity I have clean up the data and removing a long-term variability a seasonal |
| 10:49 | variation and making sure that the time series is stationary once I have done that window and as this rolling window moves towards the future um then we calculate the five indicators of Resident loss that I just mentioned uh we expect changes on those statistics either increases or decreases and to detect when time that change is start being a a warning is being a warning I then fitted and segmented regression that what it gives me is a point in time at which two linear regressions are significantly different in in the slope |
| 11:25 | so you can fit a line on the first part of the time series a line of the second and then if there is a sharp increase on the second slope that means you are getting a um an increase on the standard deviation in this case or a decrease on the autocorrelation at L one so we can do this visual checks for one pixel but it's very difficult when we have a million of Time series or or the whole planet so in order to um um not having to do this uh visual check uh what I do also is calculating the Delta what is the change of that jump between the |
| 11:57 | minimum and the maximum value for each of those statistics for now the Delta doesn't mean much but it will allow us later to make comparison comparison within um the whole planet right um the first result uh of of the study is that the genetic indicators doesn't necessarily uh agree uh with the don't necessar line with critical slowing down or speeding up theories the critical slowing down and speeding up expect an increase or decrease on variance and autocorrelation at the same time but that's the case only for a few pixes on |
| 12:28 | the data set for most pixel you find a combination of all these five indicators but is more common than not that you have an increase on on on one a decrease on the other like like the case of this pixel where for by the five metric you have a signal of resilience loss but you have an increasing standard deviation and a decreasing autocorrelation at L one towards the end of the the talk I will clarify why why that that's the case but for now uh with Delta then we can start thinking of comparison um in this case uh we constructed the |
| 12:59 | probability distribution of the Delta per biomes both in terrestrial and Marine systems uh what it allow us to do is Delta by itself doesn't tell me much but if um the the Delta statistics is on the on the ends of the distribution uh that means that particular pixel had a very big jump in autocorrelation or or or or variance with respect to all other pixels of the same biome so the number is not meaningful in itself but the rank with respect to all other pixels of the same kind might be informative of which |
| 13:30 | ones are showing the stronger signal so in this particular framework resilience loss is interpreted as pixel that are towards the end of that particular dist distribution and in the future if we will rerun the the same analysis and we track pixels that today are on the end of the distribution and move towards the centrin we can also interpret that as a resilience gain all right so just to remind you um the way we are interpreting this type of Statistics is there is resilience loss if the pixels are at the end of the |
| 14:01 | distribution so above the 95 percentile or below the 5 percentile of that particular distribution for uh the pixel with respect to the biome the number is not important what is important is what it is with respect with the others all right with that then we can look at how the map looks when we do the analysis for all the pixels of the world uh what we found is about 30% of the pixels are showing some symptoms of Resident loss uh the ecosystem that are showing the stronger signal or at Le the largest areas are Boral forest and tundra or |
| 14:32 | places that are in Arctic regions that we know are warming faster than the the rest of the planet uh we also find um signals on places that other scientists have been reporting uh resilient Lo like the south east of the Amazon forest parts of the sahil belt uh dry lands in kazakistan China India and Australia places that are water limited and where we know uh droughts and changes in the water cycle are being prominent if we look to the um Arin um data set despite we don't have good observations for the Arctic and the |
| 15:06 | Antarctic region uh we do detect a resilience loss in places that has been reported in other studies places for example where you have a boelling system like the California kurrent the vua system the Canary Islands those are places where as there is extra warming of the sea surface temperature we might disrupt the uh mixing of the water column and with that the input of nutrients to the food web so there is a observations on those places where Marine fotw might be destabilizing another big area is this blur in the center of the Atlantic |
| 15:39 | part of the Caribbean Basin that has been uh reported by other studies as a place where primary productivity is declining due to a high concentration of sing waters on the on the top layer of the of the ocean uh we are able to detect for example the big hypoxic Zone on the Gulf of Mexico but we are not able to detect the multitude of case studies of hypoxia that has been reported on the Atlantic Co of the United States and neither the coral Bing events or the coral die offs events on the great cor rft report in six 2016 and |
| 16:10 | so forth so some of the results align with what others have been done but we definitely miss some of the cases I believe due to the spal resolution or radium shifts that are not are related directly with primary productivity all right can we trust this um signals um that that was a question that I struggle with a lot and I came up with a number of tests to to see whether we can trust them the first one was a permutation test only it does it takes a a sample of pixels where we identify resilience loss reshuffle The Ordering |
| 16:42 | of events and run the signals again um if uh it was an artifact of of the data set then we should not expect differences however we do see differences in the standard deviation of the correlation for most metrics and also on the fractal Dimension meaning The Ordering of the EV matter and contains information about um proximity to theing points the second test was comparing pixels where we detect resilience loss versus pixels where we didn't and and treat them with a different method in this case was a auto regressive model that allow us to |
| 17:14 | calculate the lead value of the system as it goes over time and that lead value is a stronger piece of evidence that the system is leaving his current Basin of Attraction it doesn't tell us whether there is an alternative regime or not but it tell us that the system is getting unstable uh so up of 100 pixels for example where um we detected a resilient loss in Marine areas uh for about 50% we confirm with the second method that um uh in fact the the time series left the the the Basin of attraction or the leading value went |
| 17:44 | over one and for the other half we saw a displacement of that lead NE value towards one we saw a similar pattern but with less less strength for gross primary productivity and we did not find any differences in terrestrial ecosystem respiration data set what it help us is or it shows that with this multi method approach we can start uh testing which variables are more suitable for those some of these um early warrs exercises this is only possible because we can do it at the planet level comparing everything with everything if |
| 18:15 | you only have a couple of case studies that would be more difficult to to do that um comparison um all right so how am I doing with time abas yes you still have like 10 minutes I think perfect um so I have some concluding remarks for now uh but um before I I close up I would like to get a little bit into the problem of whether the critical is Lo or speing about a contradiction um experimental work has shown that with an ecosystem say a community on a petri dish is Disturbed with more than one drivers you are increasing pH changing temperature and |
| 18:56 | salinity at the same time uh the metrics of of proximity to theing point might generate this apparently contradictory signals uh so that we know experimentally happens when there is multiple drivers uh at play um in addition to that even if you only have one driver the same driver can have impacts changing slowly so approaching the vcation from a from a slow driver type of perspective or changing the variability and these two things can happen at the same time uh because of that you can also expect that the same |
| 19:27 | driver affecting the system in different ways might create these contradictory signals to test a little bit of that idea uh I did test some um random Forest some some some apply machine learning approaches to try to explain what is explaining the detection of resilience so here on the left you have a logistic regressions on the top for terrestrial ecosystems on the bottom for marine ones and on the right you have uh the same models but with a random Forest approach uh let me then zoom in and let's focus |
| 19:58 | focus on on Gross primary productivity uh the logistic Recreation what it says is after controlling for geography and for biome type so we know for example that Tundra Boris are more likely to uh show early warning signals or symptoms of resilient loss after that we are testing whether if variables related to climate like temperature or precipitation or changes in land cover explain additionally um statistically speaking the probability or the ODS of detecting symptoms of resilience so here the the model is comparing pixels where |
| 20:31 | there was early War indication versus pixels where there was not uh what we see is that the linear Trend or the mean of these particular pixels is important but it's also important the changes on on the variation in terms of the long-term variation the fast oscillations or the annual cycles and for some of those pixels um for example precipitation the variability of precipitation on drops are changing faster that the the mean average change of of of of precipitation itself uh so it trace the FL that um we might |
| 21:01 | approaching this uh critical transition bya changes on on variability rather than uh the typical bifurcation approach where critical slowing down is very successful um in the random Forest we see about the same uh type of evidence when me temperature is the stronger predictor of um resilient loss for example for gross primary productivity but the next three are all related with changes in variability in temperature at different time scales uh and the same for precipitation and the variability of precipitation in |
| 21:32 | different time scales so this uh help us um reinforce that intuition that uh the same drivers are affecting ecosystems in multiple ways are approaching tiing by the slow variables or changing variability or that multiple drivers are affecting the same systems at the same time as I discussed before uh a lot of the patterns detected here align with work that other people are doing playing around with different data set like ndbi uh VOD which is vegetation Optical dep and so forth so that uh kind of reassures um that we are getting um a |
| 22:08 | robust or consistent signal at least qualitatively speaking with uh what other groups are doing I would like to emphasize that the results here I believe are an an underestimation uh the reason is the limitations of that resolution remind let me remind you that we did not detect things on uh cases of hypoxia because they tend to happen in a smaller time sorry special scale at which we have pixels for uh and we don't have um ground truth either for for this data set or we do have some ground TR for example the the regim ship database |
| 22:42 | provides a series of annotated cases of regime shifts around the world but it's bias so it will induce bias on the models uh for places that are well studied like the global nth or Coastal hypoxia where we have hundreds of cases and then um it will allow us to see things things on the global south or things that are not so well um documented despite those limitations I believe the results allow us to do some um qualitative uh understanding of where things are getting better or worse we can for example rank countries by the |
| 23:14 | area of uh of their ecosystems that are showing symptoms of resilience loss uh of course countries with higher area will also show up on on the ranks of Russia Canada and the states are are the top three countries but if we rearrange it by the proportion of territory then we start seeing that ecosystem that have a rich amount of ecosystems that are um desertic of water limited system like kakistan Mexico Mongolia also start start leading the the ranking the same with Australia on the B panel you see the countries ranked by the diversity of |
| 23:47 | ecosystems that they have so we can do the analysis at the level of biome which is where I presented but uh um we can also look at the diversity of ecosystems and again we rearrange by proportion of territories highly biodiverse countries start leading the list like India Brazil Colombia Australia Canada so with that um I would like to close I leave some some space for questions and um yeah very thanks very much for your attention and and your interest on this seminar let me stop J yeah thanks Juan for this really interesting uh talk so just a |
| 24:25 | quick reminder to everyone if you have question please pause them into the Q and a box and we'll address them during the discussion session maybe in more detail but we have two question one so I think we have like four minutes before the talk of helmo so maybe you can already answered this question one question from Isabel repa so um she's asking if is citizen science use to get data information on resilience laws of ecosystems or to Comm unicate about it uh thanks Isabel for the question right now we are not using citiz and |
| 25:04 | science uh we have some plans of using the regim ship database and people can report cases there but um we don't have many cases reported but but not scientists on that particular resource uh I know of work that they're trying to train for example um text text type of models to identify symptoms of resilience loss or newspapers or things like that so it won't be citizen science but it will be train on data produced by citizens journalists and other sources yeah and there is another question if from case alerson so if an |
| 25:39 | existing ecosystem in one of these pixels is destroyed and replaced with a new one that is more resilient to climate variability as measured with your indicators so um yeah does this uh qualify as resilience loss or gain well the resilient losses in this particular study is not about the ecosystem type it's about the function that the ecosystem is performing so the when I use the gross primary productivity data set what I'm actually measuring is the amount of carbon that the consistant captures over time uh so |
| 26:15 | is that that is the function that is uh being becoming unstable or yeah destabilized if we change from forest to Savana the two type of ecosystem will perform and that function differently and uh which one is more desirable will depend of of you know the the managers the community the countries that um govern that particular ecosystem but I wouldn't say that the one type of system is more resilient than the other uh both are resilient depending on which which type of uh disturbance you're talking about and um in this particular case as |
| 26:46 | a clarification I was talking about resilience of primary productivity which is a function of those ecosystems but it could be different if we measure resilience of the water cycle or some other function that ecosystems perform yeah thanks one I think we will address the other question during the discussion session thank you so um so please don't forget to uh direct the question to the relevant speaker because to not like uh losing between who should answer your questions so um and it's now my pleasure |
| 27:19 | to introduce our next speaker uh helmot heland so helmot is a professor for Plankton Ecology of The Institute for chemistry and biology of the marine environment at the University of aldenberg he is very interested in the mechanism that constrain and alter biodiversity in a VAR a variety of ecosystems so he likes to think about the complex nature of biodiversity change and ecological stability he is trained as an experimental ecologist and he has focused more on research scienes and uh data analysis in recent years and |
| 27:58 | since 2017 he is founding director of The helmholtz Institute for functional Marine biodiversity at the University of aldenberg and which is like a collaboration between the university and the Alfred vegna Institute uh helmold Center for Polar and Marine research and their focus here is on interdisciplinary projects on Marine biodiversity and conservations today helmot will explore whether there are threshold for biodiversity change or not so thanks for being here helmot and the floor is yours thank you AAS for introducing me |
| 28:42 | thanks Hanan for a fantastic talk um I will look at a similar question but from a slightly different angle and that is mainly from the angle of biodiversity not so much uh the emergent processes or um functions as associated with them um I'm standing here as an author but this is uh standing on the shoulders of many people who worked on this and I will introduce some of them at each section I also gave a list of the main things I'm alluding to in the chat so if you want to know more um feel free to to check |
| 29:19 | these papers um so we are all aware that biodiversity has been changing in recent decades at an extent that has not been observed in um uh prehuman times and uh the 2019 Global assessment on biodiversity by IIs has told us that some of these groups have a dramatic amount of u species percentages being uh threatened or or even extinct and um um we have this this Global phenomenon of um um biodiversity change going on um in in almost all types of ecosystems on Earth the dire truth is that this is not stopping uh anytime soon um Dy hup who |
| 30:15 | is pictured here uh published a paper early this year where they took data from more than 30 3,000 marine species and try to exp extrapolate what is happening under different climate scenarios um they looked at other factors uh in addition to temperature um but actually they only took climate as a driver so this is the most optimistic view you can get because it's actually excluding all other changes that might be ongoing as well and depending on how optimistic you are about our climate Pathways uh in the |
| 30:56 | worst case uh 33% of the ocean will change more than 50% of its species inventory half of the ocean uh will uh half of the species will lose at least half of their habitat size and for almost all equatorial species 88% people uh um find U distribution gaps um um uh around the Equator so this will be massive and this has led to uh claims that we are transgressing a threshold uh in how much biodiversity change this planet can endure um and uh Yan Rock's famous uh figure has just been updated a few weeks ago um uh with um uh a very similar |
| 31:53 | message that biodiversity loss has actually um transgressing uh planetary boundaries for for um um our planet so this um opens the question how how do we deal with that how do we uh incorporate that into our management and the big thing is that management is not happening at planetary scales it's actually happening at local to Regional scales so how do we Implement biodiversity change or a threshold for biodiversity change into management and this is the three aspects I would like to quickly address in this |
| 32:34 | talk um are there thresholds of biodiversity loss is there thresholds of pressures on biodiversity and um what does that do with our Target setting and uh and management the first two questions come from a very conceptual issue if we think about thresholds and biodiversity change we actually have two ways of thinking about them and they are not independent of each other they are actually quite closely linked but I think it's important to disentangle these two because they have different they cast different views on the problem in the |
| 33:13 | upper graph we have something that pressures biodiversity and we observe at some critical level of pressure a change in the response of biodiversity so uh we see a disproportional erosion or change in biodiversity um at this uh threshold level of pressure so what we see on the y- axis you could call a tipping behavior and what we see on the x-axis you could call a threshold in the blue example on the bottom it's actually the biodiversity change itself which is on the x-axis and we can have a certain amount of bod |
| 33:57 | change without changing the functionality uh the carbon capture the primary production and so on uh but then after a certain amount of biodiversity change we will um lose type part of this functionality and uh we uh might even lose the entire functionality um in that system so starting with the lower example quickly um and as we have only uh 25 minutes or so uh I will um go over this uh a little bit quicker the lower example would be ideal for a management purpose because we would actually be able to define a safe |
| 34:44 | operating space that is far away from a threshold where we actually have a critical transition in the functionality of the system and if you look at the first generations of the so-called biodiversity ecosystem functioning research we might actually find patterns that uh look very much the same in contrast to My Graph the x-axis is reversed here so it's biodiversity on the x-axis not biodiversity loss so you have to quickly mirror the graph uh for comparison but in this uh summary of several uh hundred experiments uh by |
| 35:27 | Brad Cardinal um more than a decade ago actually a very large proportion of these studies showed exactly this type of relationship with an ASM toote and then after losing a certain amount of species you all of a sudden see a decrease in biomass production or decomposition or nutrient uptake or whatever function you're thinking of however this first generation of uh BF experiments has been criticized very much um because it actually looks at one aspect of biodiversity normally richness and one single function at a |
| 36:09 | time and people have very early suggested that actually if you take the multitude of functions this community provides to the ecosystem into account you're not seeing this type of saturation function anymore but you are seeing a very a much more linear relationship and this has been um termed to or coin the the term has been coined the biodiversity multifunctional relationship and that often appears grle uh almost linear um one of the best examples is a grassland study by uh numerous colleagues uh the y u biodiversity manipulation experiment |
| 36:56 | that runs for two DEC Dees now and uh there are up to 80 different functions have been measured and if you look at um the slope of the diversity effect it actually becomes stronger and more linear the more functions you are actually adding um to the analysis so it actually means that biodiversity loss that is safe is something that is potentially an illusion um and uh um uh the the idea that there is some amount of biodiversity that is redundant um uh disappears as soon as you um look at the multitude of processes in the in the |
| 37:43 | ecosystem however and that is uh more an ethical than or philosophical remark than than a natural science remark even if there would be something like a safe operating space who is saying what is redundant I for example can confess that I have a laugh for orchids even though I'm a marine ecologist uh and I don't want them to go extinct um so I think even the idea that there is an amount of biodiversity loss that is tolerable is ethically wrong um and brings us on a very strange argument when we think |
| 38:26 | about what type of um impacts on natural ecosystems do we still tolerate uh if we think we as humans have the right to say which biodiversity is dispensable or not so the remainder of the talk will be on the red example which perhaps is the more realistic one uh that we have a pressure strength that induces some kind of biodiversity loss that is much larger than um um U than than than you would get get from from a proportional change and um I will look at this from um three viewpoints one is what is actually data |
| 39:18 | telling us in this respect um would we actually see the biodiversity change in time with the pressure and what is about delayed responses and what would be our ability to detect these disproportional changes if I think about systems that uh might have a pressure level where they actually have a disproportional biodiversity change I always think about spe uh ecosystems that depend on Foundation species who are showed the cor Reef example you can also think about seagrass Meadows for example in the marine environment where the |
| 40:04 | presence of a a species or a very closely related group of species is actually defining the ecosystem and the absence of this Foundation species obviously changes biodiversity so you will observe all um um or more gradual changes in the composition as long as the Foundation species exist and if they don't exist anymore then you will see a rapid change also here of course we are already making a judgment because uh in this nice graph by EMT Duffy um which which I've fully subscribed to uh that this is one of the changes we are |
| 40:46 | observing in benic habitats for example we are of course putting the diversity of all the Plankton and I'm a plon plankton ecologist so I have to make this remark um as like one gray dot but in in reality of course it's many species so if you would just count species you might even increase in biodiversity in the suchal system but that is only the smaller site here if you look at data of how biodiversity changes this is from two coastal areas that have been undergoing a lot of environmental change over the |
| 41:20 | last uh decades um and if you think about how the composition changes over time with temporal distance what we are mainly seeing especially in the long data series is radal change the similarity increases with temporal distance it's not increasing Suddenly It's actually increasing slowly and actually this might be one of the major functional consequences of biodiversity in itself the incredible number of feedback mechanisms that a high biodiversity insurers actually might smoo transitions uh because exactly of that of these many |
| 42:06 | feedback mechanisms that occur this is an example Again by D hup um uh where we um looked at uh turnover in grasslands if you take or if you change the environment drastically so this is a glob experiment uh 50 sides worldwide uh same treatments and we look at the compositional change when you take away herbivory and herbivory in grasslands is one of the most important influences you can think of and if you take away nutrient limitation and actually what you see is no rapid change it's actually a slow turnover and actually the treatments are |
| 42:52 | less important than how much additional diversity you have in the meta community so how high your beta diversity is it's actually the existence of additional information around the target plot that is constraining biodiversity more than the change in this environment so again we don't have an approp response to an aupt change I've recently started interacting with paleoecologists and this is a paper that came out um early this no late last year it's uh led by tonu here to the left uh PhD student at the University of |
| 43:37 | Raymond and she looked at core data over the last uh Del glaciation period in the Atlantic Ocean and what you see on the lower graph uh to the left is the global temperature change over this time series you see the Dr atic uh increase in temperature uh from 177,000 years before now to um um 12 uh 10,000 years before now and then on the upper uh graph you see the change in community composition and again it's a gradual change and it's actually continuing after the temperature change has ceased um and this um made us think |
| 44:27 | about what what is the synchrony between the change in the environment and the change in the composition and we started with a thought experiment this is mainly Work by Lucy konski um posto here in ook and uh we collaborated with uh presentent ontiveros for a neutral model that I will show in a second so think about we are doing something really bad to the environment putting a toxin in every species is negatively affected however we will not see immediate extinctions uh unless we are putting so much toxin in that that everything is |
| 45:10 | killed but that is like trivial um but in a system that deter deteriorates what we do is we turn population growth rates negative but before the population actually goes extinct there will be um a time delay unfortunately and that is something that we really have to communicate better to managers is we also have the the the opposite problem if we do something good to the environment like in the middle panel all species might profit but before we see new species arising there is a colon colonization delay but the real interesting thinking |
| 45:50 | starts when you think about a change in the environment that is not per se good or bad for all species species like warming it will attract some species that more like it warm and it will deter some species that more like it cold you could argue that in many systems at least those that have not a very large isolation um that are not very isolated sorry um colonization might actually be faster than Extinction so uh you might actually only need a few uh individuals to um um to to increase diversity whereas Extinction depth is something that is |
| 46:37 | already known in the ecological literature mainly from the fragmentation literature that species of course not immediately going extinct but they will slowly be replaced or uh um displaced by by um new colonizers it's also true if if you monitor these types of of systems that all species that go extinct must have colonized at some point so they either must have been there so there's actually uh a as I think underexplored um um net imbalance between colonization and Extinction here so we looked at European freshwater fish data uh it's |
| 47:23 | the Riff fish time database by uh Lucy comter and others freshwater fish are meant to be extremely vulnerable to environmental change um freshwater systems in total uh are accounted to be very vulnerable and what we see is all kinds of Trends each of the gray lines is one of these more than 3,000 time series um but on average we see an increase in richness so we have this group of species that we think is extremely vulnerable but it's actually in increasing in the number of species if you do a classical null model |
| 48:01 | it predicts a zero slope as we would expect however it completely ignores autocorrelation in the sense that I depicted in the previous uh paragraph if you instead do a neutral model that includes this autocorrelation we actually see an increase in the number of species again so even though this neutral model has no feature that would allow species richness to change it's actually changing simply because colonization is faster than Extinction and um this um net imbalance in uh colonization Extinction actually um um makes our task to |
| 48:46 | identify this pressure strength even um uh more difficult we don't know how many of them uh um of these communities actually show this type of of uh disproportional change and the trigger might actually be much earlier than we actually observe in our change because the delay in the response of biodiversity um uh seems to be um um a very strong uh signal the third aspect is if thresholds exists in order to use them for management we would actually need to be able to predict them um measuring them our posterior |
| 49:37 | doesn't help us in our management if we realize that we now just transgress the threshold uh we are not having a chance to go back and say uh we should we should not have done that um so this is a little bit um a the history behind this paper um that uh we published in this team here um um um where we uh looked at what what kind of information can help us to um uh identify these thresholds many people have used time series but time Series have the problem that if uh the threshold is still far away they of |
| 50:20 | course don't show any signal of the threshold but we don't know how close they actually might be so we thought about how about experiments not any kind of artificial experiment but experiments that took whole communities suped them to change and looked at a response of one of their emerging properties so each experiment consists of a control and a treatment the treatment is any kind of pressure warming uh ufic whatever and they look at a response variable and they you can use these information to create a response |
| 50:59 | magnitude and this response magnitude because we also have information on the variance uh can be uh added with a measure of confidence and uh by putting many experiments together you could actually think about one one option is that there is a gral increase in response magnitude when you have stronger pressure however if there is a threshold where um the pressure actually leads to a new response magnitude then you should see something like uh a certain smaller response magnitude at low pressure because even the most sensitive systems |
| 51:43 | are below their threshold then some actually flip into something new a new type of response and at the highest threshold level you would even put the most robot robust systems to the um to the brink this is acknowledging that for each of these experiments there is a different threshold of course but you could at least say when when do these types of responses kick in and we did that for four and a half thousand experiments from 36 meter analysis uh it's important to acknowledge that for each of these meter |
| 52:23 | analysis there was only one thing on the xaxis like temperature in degrees celsius warming and one thing on the y-axis like biomass so it's not merging everything to everything um so we have 36 cases unique combinations of pressures and responses and we looked at whether the pressure strength impacted um the response magnitude independent of whether it was grle or disproportional we checked whether we actually find a new type of responses along the pressure gradient um so is there an increase in variance and we |
| 53:04 | tested whether we find multimodal uh response magnitudes uh we found that in most cases the pressure strengths of course had an impact that is what we expected we had eight out 36 meta analysis where we found this variance increase however in two it was actually a variance decrease that we found so not what we expected and in three more it only occured at the highest ever applied pressure strength in ecological literature we had no uh multimodal distribution of response types as you would expect if the um if it's a very classical case of |
| 53:52 | regime shifts one thing that has been misread into this paper even though we explicitly State at least twice um is that we say there's no thresholds this is not what the paper is about it is about detecting thresholds there might be a lot of thresholds but if we can't detect them then we can't um use them for uh uh conceiving future management issues another aspect that has been published here by be spake and John Chase and others is that uh they uh uh said that uh we might have underestimated our |
| 54:38 | ability to predict these thresholds because we used Lo response ratios so we had Lo transformed response magnitudes however we had had already done this in our scenarios we did test of course whether our um three tests would actually detect threshold type of uh um uh increases in in um um uh in in uh responses and uh they actually were without noise they were Flawless so they we we were able to detect them when they are there however and that comes back to something Hanan has already said uh already small amounts of noise in the |
| 55:28 | simulations so where we Jitter around the data a little bit um actually makes us very um or or lowers the the power of these tests very very strongly uh independent of the lock transformation so um uh that made us sorry for intervention but we have already crossed the yeah it's my second last slide yes um so what I definitely would like to convey and stop with is uh to say that um uh we don't want another locked in debate whether thresholds are present and and and uh how real they are and so on I think that is that is not the case |
| 56:17 | here I think the important question is can we widen the context in which we discuss and the perspectives from which we discuss to realize that um many of these um aspects we we try to convey are actually more similar than they are separate and the most important one is that um the um our emphasis on biodiversity change is something which is occurring slowly and gradually is not showing that we are safe by contrast I think the threshold debate sometimes leads us to assume that within the safe operating space we are |
| 57:02 | not in danger of anything but we are actually losing a lot biodiversity already at very low pressure strength and I think uh therefore uh I'm critical in um how we can use um uh this type of threshold approach to manage biodiversity well there is a number of things we can do um actually it's not so far away from what also Kanan has presented uh in how to make us better estimate the uh our ability to base our biodiversity Management on such uh threshold approaches thanks yeah thanks thanks helmot for this excellent talk which I think uh |
| 57:53 | nicely complement what Juan also presented so we are now taking questions for uh both speakers so please don't hesitate to pose your question and to kick off the discussion session and while we wait for more question popping up so I would like to ask um a first question I mean to maybe both of you um so both of your talks uh touch on that Dynamic interplay between like biodiversity resilience and ecological stability and given the like ongoing crisis of climate that we have today and increasing human pressure so how do you |
| 58:40 | unvision the future of our Global ecosystems like specifically for example the most vulnerable ones so I'm just wondering if um there will be like new imer of new ecosystems that can also like function in place of our previous ecosystems so I'm I'm really happy to hear your thoughts about or your answer Helm would you like to go first ah that is a very broad question um I think one of the things that often is kind of um um um or one thing that that that I think we need to consider is when we think |
| 59:35 | about stability do we think about the stability of an emerging property or a um or the biodiversity itself and I think these are two different things and and sometimes we kind of are very clumsily um assuming that they are very closely related but they they aren't um if you change the environment to put it very acute the only way to keep your function is to change the biodiversity that Pro promotes this function because you will shift species out of their Optimum and if you want to harvest something or if you want to have carbon |
| 1:00:16 | capture or something like that this will only be done by other species that are better adapted to these new conditions so if you are a harvest manager or a functional manager you would go for let biodiversity change as much as possible to be the optimal biodiversity for these new conditions if you want to protect the roseros this will not help um so if if you are a conservation biologist and you want to um uh keep a certain species or a certain species composition or even an ecosystem um then then uh you uh have to |
| 1:00:58 | be aware that under different conditions this ecosystem will perform differently so the function will change um so there is actually um um a a relationship between functional and compositional um changes uh in the environment or in these ecosystems and I think one of the interesting um discussions that we we are not really uh into very much at least what I'm reading um is how how do we bring these two things together how how how do we evaluate shifts in biodiversity where the shifts in the stability of a certain |
| 1:01:44 | function for example yeah I mean I do agree because it's always like like separated study I for example structural or composition of the species but but we have also the functional Spees that's also like providing this ecosystem Services also for for people so so sometimes the focus is is is changed so and we should I think we need to be more precise in in in the scientific Community when for example we introduce results or whatever as threshold or as like level for for which function for which species for |
| 1:02:22 | example so one do you have any no just to complement helmont's answer I think there is a a strong line of research developing on Noel ecosystems uh if you remember his first or second slide we're expecting uh populations in the marine environment to move uh towards the PO to towards CER areas inter terrestial ecosystem we're expecting exactly the same going up uh in higher latitudes or higher in the mountains so there will be Noel uh communities as we go into into the centry and and the the scenarios of |
| 1:02:55 | changing climate precipitation and and factors that determine how these communities uh are assembled um that been said uh not all those changes have to be bad and not all all those changes have to compromise ecological functions the as helmo say I think that there is a a a wide open door there to do some research on how do we distinguish uh where some functionals that we care about are being compromised versus how um yeah communities are are transforming and following adapting to these new conditions that we |
| 1:03:26 | are putting the planet on yeah thank you I think I will take some question also from uh the Q&A box um I have other question also but I will delay them in later and the question is uh to to Juan so um it's from n Davidson so are you watch this data changing how do you feel and how do you cope so he's referring to the inner work required to face the outer reality so how how do I feel sometimes with this particular data set is is not a a when I was doing my PhD I was putting together the regime shift database which |
| 1:04:12 | is a that was more emotionally Bome because he was just reading papers about bad news of how ecosystems are changing and and and being degradated but I think uh working with these data sets and being able to point out where problems are getting worse or or making these Maps uh put as a scientist on a privileged position of a communicating where actions might be more um useful uh some of my my projects with my colleagues here at SRC are trying to translate this for example to companies uh or or or to actors that might be able |
| 1:04:45 | to do something about it so I feel this this paper that I presented today was my coid paper and as we were in coid all feeling uh depressed about being en closed and and being the situation around us I was thinking often on the doctors that facing the crisis just on the borderline and and the people who came out with the vaccine and the people who came out with the science that allow us today to be on some sort of back to normal so I feel some sometimes that I'm I'm privileged in doing some some of the |
| 1:05:12 | science and trying to understand where actions might might have a bigger difference in the future so I think it's a it might be um it might you feel depressed at points but uh it also give me hope that my my work might be um useful yeah thank you there is another question from Max TR so is there a natural Dynamic of ecosystem resilience uh not being linked to human activities indeed H it's also for me yes I think so Max uh if we imagine a parallel universe where there is another Earth with humans on it uh ecosystems will still have some |
| 1:05:50 | Dynamics some basins of Attraction and some perturbances uh uh and then they will be facing um issues issues in that par par World resilience is also relevant so without humans we can think of resilience but uh in the planet we live today humans are a very strong force of changing ecosystems and it's really hard to separate them so in that sense in this planet we should always keep them in um I'm going in the order of the question actually so there is uh OTA Clen who is asking do the results allow to make conclusions regarding the |
| 1:06:26 | proposed tipping elements of boreal forest so yeah uh conclusions about the Tipping points per set not necessarily it it identify the areas where things might be going on that direction I think other groups have been identifying thresholds of temperature for boreal forest and and and Tundra and also the temperature at which uh the ability of forest to sequest carbon change so when we have the from uh source to um sync to source of carbon that has been work done also with a very similar data set that I use here and also some observational |
| 1:07:05 | work in the Amazon where this has already been reported so from the satellite you still see threes and you see still the same diversity but their ability to sequester car carbon is what have changed yeah yeah thank you Juan and there is another of course how long is um is long enough to detect the Tipping Point to detect a Tipping Point so the I just want to clarify that I'm not detecting tipping points here I'm detecting uh signals of residance loss which is a symptom where a Tipping Point might be close or not going back to the analogy |
| 1:07:38 | of of coid is like I'm detecting that you tested positive for coid it doesn't mean you are going to die it doesn't mean you're going to get long-term symptoms it just means you're presenting some symptoms so it should be interpreted that way uh one can start thinking of fit in some of the CES that helmo was showing on his statistical approach uh to see whether the pixels that show strong resilient lws are related with certain points of climate or temperature or precipitation change or things like that um I'm not sure this |
| 1:08:08 | time series are long enough to do that is is um because the primary productivity changed very slowly over time we were looking at some of their functions probably uh we could think of of something like that like a pollination or or something that happens faster and a smaller scale we probably could detect something like that but with two decades is quite difficult and if we Adventure into more climate typ in ele elements definitely this is not long enough to detect uh something like that at least from the statistical point of |
| 1:08:38 | view yeah yeah thank you and there are there is also another question that the loss of resilience is not necessarily linked to uh the presence of tiing points so Behavior so did you also test somehow if assist system has hm marks of Tipping Point systems I don't know if that question is uh clear so it's uh Jim Van Belson who is asking if thew the Los that the loss of resilience is not necessarily linked to the presence of Tipping Point Behavior so did you also test somehow if a system has H marks of Tipping Point |
| 1:09:23 | systems oh we or there is indicator of like this Tipping Point I think yeah so so far I did not test for the existence of the Tipping Point itself I as Helmut I believe that it's really hard to statistically get a a good answer to to that that being said we have some some uh people working on that problem um as as we speak uh but um rather than the if when I was a kid I really like to destroy things and my mom didn't like me because of that and then when I look at a regime shift I often think on how many different ways you can |
| 1:09:56 | tip it over how how many ways can you screw it up so you think of the Amazon uh you can you have a Tipping Point already quantifi which is between three and four degrees Celsius but before we get to that we probably going to cross Tipping Point fragmentations which is proposed at 40% we're already at 25 or something like that you can also cross tipping points in terms of fire frequency or change the the moisture Recycling and then the amount of precipitation that falls in so it's it's not just one Tipping Point is when you |
| 1:10:24 | have multiple ways of tipping you might reach one or another uh earlier and reaching one might increase uh your probability of uh crossing the other so I don't think the numbers in itself are very informative what is informative is to know those different ways and acting in a or managing systems thinking plurally on how to avoid Crossing them uh the maps I I I show is just showing me uh where those actions might be more useful yeah thanks Han yeah like great so we have also a question for you helmot um so from case alerson so she's |
| 1:11:04 | asking this uh year the convention on biodiversity met in moral and acknowledg that none of their past issue biodiversity targets for 2020 were achieved uh they then created a new framework that is due to be achieved by 2030 and if any more more ambitious than the isue targets were please comment on any changes between then and now that might lead us to believe that new biodiversity framework may or may not be any more successful than the a targets where yeah so uh fantastic question Keith um and um first of all |
| 1:11:51 | um uh the um Global biodiversity framework has taken some lessons from the iich process had has tried to be more concise um and clearer in the demands for um the change necessary um we are just finalizing a um paper at the moment where we look at how well do the global biodiversity framework targets reflect uh what is being discussed about biodiversity change in the literature uh in the scientific literature it actually it does very well but that is already coming at the problem it's actually not a failed biodiversity framework or |
| 1:12:44 | our inability to provide the scientific basis for making decision it's the inability to make decisions so the a major hinder in a more positive effect uh on biodiversity is that we are not putting these targets into action and uh I think that is uh something that um also took me a while to realize we are doing a national assessment of biodiversity status and Trends and solutions here for Germany at the moment that shall be published next year and uh we realized that um the the biggest um hinder why for example area |
| 1:13:40 | protection is not uh effective is that there is actually almost no measure in the Marine systems for example we have a number of marine protected areas that actually exclude not a single use and uh it's not very surprising that it's actually not any different neither in it abiotic condition nor in its biodiversity through the surrounding because there's actually nothing happening so we are actually mainly facing uh the problem of um um of of using what is out there as information and the targets that have |
| 1:14:23 | been proposed in a good way on the other hand and to be a little bit more optimistic we might actually have reached um um a a status where where uh this awareness becomes um um more or broader um in the uh Global political community so the fact that we managed not we as me but we as humans um um manage to get an agreement on biodiversity Beyond national jurisdiction despite all the political conflicts that we have and even a war in Europe and and so on is actually something that is U can be taken as a um as a step |
| 1:15:20 | forward uh 2030 is tomorrow and uh that is um perhaps one of the uh biggest problems that there will not be so much happening even if you would reduce our impacts dramatically now and uh create um large areas where U nature can develop uh we would not see the positive outcome of these measures um um until 2030 uh some of these species we are talking about will not even had a single generation since until then um so if you if you do only very fundamental population biology you will see that this is not this will not happen um so I think we |
| 1:16:17 | also have to think about the targets uh in a more or the time frame for our Targets in a more realistic way um than than we sometimes do and I'm fully aware that in most democracies we have like four to five years election cycles and that is often what drives policies but that is not helpful for um for for biodiversity yeah I think I I do agree with you helmot regarding the timeline because this is laking really from this report of CBD or other International report and when we are mentioning for example in some Target like Target to |
| 1:16:59 | Restoration of 30% so we are talking about restoration but we're not talking about the timeline to restore this like ecosystems Etc that we put the target to 2030 but we need more time to restore this like area or this EOS or this species or whatever and when we talk also about like in Target to about like degradated land then there is also different degree of gradation and uh which which type of degradation do we need so I think we need to be more um precise in this like uh targets or or or in in putting this targets like that's |
| 1:17:37 | why it's like difficult to translate them to to actions for example because uh we don't know and each ecosystem is unique in its characteristic given its characteristic and all these complex interactions in in biodiversity with the environment and these kind of things so yeah I do agree uh that we like that in in in this uh CBT report if I may add one thing we often and Global biodiversity framework and iich has done this as well we often think about the things that we protect like this 30% protection the |
| 1:18:15 | 10% no use protection this will not save biodiversity I think we also have to think about the other 90% uh and the way we can't like uh uh completely destroy 90% and think that the other 10% will uh uh make for an healthy planet that is that is not working we really have to think about reducing pressures um um in um um in in order to to uh uh bend the curve in in the biodiversity development yeah I I do agree also I mean for example example like human modified lands also are overlooked in this like reports so for |
| 1:19:00 | example where we can regenerate also functions and services and biodiversity and which can complement also the intact area because now we have almost like a half of the global land like transforming to agricultural land and human modified lands and this is also like missing in this um in this like reports um there is I think another question for you helmot so it's from n Davidson um uh so helmont's response about awareness in conjunction with the biodiversity convention seems to allude to social tipping PS so what do the |
| 1:19:45 | speakers think about social tipping points and the time left on the clock of the world yeah I had hoped you had uh read Nile's other question which is much easier to answer um anyway we have just like two minutes if you yeah answer so uh uh to D's other other question on a very abstract level we actually know who the winners and losers are uh but we can't predicted for a certain ecosystem and that is again the problem that management is often at ecosystem scale and not at um uh global scale where we have perhaps this this |
| 1:20:29 | type of knowledge uh I'm not a social scientist um I can't really say something about social tipping points I would ask myself a similar question whether there is something where um the ability of a society uh for example to absorb change is something that um that um has a gredle or less gredle or even tipping type of of behavior um I I'm I'm simply not read into that literature to really make a make a true answer on on this question but uh I think it's a very um uh important one I don't know Juan if you can also |
| 1:21:21 | you can also jump in and if you have any feedback on this uh comments no just just a complement I think the literature of socialy pinpoints is taking taking off uh with some people who believe or disbelief on them and I think that discussion is a healthy one to have regardless of whether they are there or not we do need to uh promote an energy transition and a number of transformations in our societies to to tackle the the problem that we have ahead all of those require a critical mass of of people joining hands together |
| 1:21:51 | and and making things happen whether in theend energy or in natural resource use or things like that uh so I think whether there are social tipping points or not is is more of an academic discussion is more on the hands on how do we make those actions meaningful actions happen by meaning meaningful actors and how do we make sure we don't leave people behind so so some of those values of the sustainable development agenda that becomes relevant here I think we can uh take one last question I don't know to who is directed |
| 1:22:24 | but you can yeah you can you can answer both so if we can prove that a threshold exist but that we cannot predict them with sufficient uh accuracy to influence management can we make an argument for uh precautionary approach based on the analysis of the consequences if a threshold is crossed so this is Philip uh Churchman who is asking this question I think so I think uh uh the science help us even if we don't understand what's the exactly number of cigarettes that give you cancer we can still promote policies that prevent the |
| 1:23:02 | smoking on on young people or that prevent bad things of of happening so I I do think that uh we don't need the the quantification of the number uh to be able to design meaningful policies that allow us to to manage U our ecosystems in a in a way that allows our future generations to enjoy the same services that we enjoy today yeah um I just can amand that if so we we played around a little bit with like very uh core scenarios about uh assuming a threshold at the wrong place assuming a thresold exist when it's not existing and so on and so |
| 1:23:41 | on and what this means for the the the costs of activities um in that norre paper that I put out in the chat uh so if you want to um read that as I said it's a very very simplified toy model uh I will actually be at resilience Center I hope in winter um and uh I already discussed with yon that we might actually one of the things we might do is actually think about to making it a little bit more tangible um at that point it was more for our own information than from for for anyone else um so if you uh if you're interested in that uh |
| 1:24:26 | we might check there for very very small glimpse into what what this means yeah thank you so much for um your great answers and uh talks uh during this event I think we have just like few minutes remaining so we need to conclude this event so we want to give a virtual Round of Applause and thank our speakers and again for for their excellent presentation and for answering our questions so and many thank to to all of you still online for your attention and participation today um just a reminder that we have uh |
| 1:25:11 | several events coming uh up on economic of tipping points and the Tipping of methan and also if you would be interested to get involved in the Tipping Point model in inter comparision project so please uh see the links in the chat and please sign up or express your uh interested and thank you again and have a great morning afternoon and evening wherever you are thank you so much ju and helmot also for your very nice talk thanks everyone have a great day bye byebye five |