Joshua Ginsberg
All right. The last thing is, we are doing our summer ecology camp again. We're doing more of them this summer. Now the pitch I'd like to introduce our own Dr, Evan gorra. Evan has been here for two years, but he's been doing this kind of work for many more years than that undergraduate at University of Pittsburgh, get that right, yes, and University of Louisville for his PhD, he's a fellow at the Smithsonian Tropical Research Institute, another federally funded scientific endeavor, and has been looking at the question of what kills the really big trees, and I will say no more, other than it turns out to be both a really interesting question, a really important question, and not an easy one to answer. So I have heard Evan talk before, so I know it will be a wonderful talk, and I'd like to thank him for entertaining us this evening.
Evan Gora
Well, thank you, Josh for the wonderful introduction, and thank you everyone for being here, everybody in person, and also our audience online. Really appreciate everyone taking time out of your Friday evenings to come and learn about some science. And in particular, today, you're going to learn about a project we've been developing called gigante. This is. Anybody here know what the word gigante means? Can you guess? Yeah, giant. Yeah. So we picked gigante as the name for this project for three reasons. One, if you speak English, Spanish or Portuguese, it's pretty obviously giant in Spanish and Portuguese, it's just actually how you spell it. And then if you speak English, you can kind of infer, and our team mostly speaks those three languages.
The second is, the project itself is giant. There are more than 30 collaborators across more than 10 countries, and rolling this out in all the places it operates has been a huge endeavor. And the third reason is just very, very on the nose. We're studying giant trees. So naming a project giant seemed seemed pretty Pro. Now, to get us started, though, before we get into gigante itself, I want to set the stage and first take us to tropical forests. So this is a drone video right here.
This drone is about 1000 feet up over the Amazon rainforest in northeastern Peru. I took this video in 2019 I think the first thing that hits you looking at this is just the scale. I mean, this is enormous. This river right here. Does anybody know what river this might be, Northeastern peril? This is, well, it has multiple names, but it's what will become the Amazon River as it flows down right before it reaches the Napo.
And you can see this is an island inside of a massive river, a cumulative River, is probably around two miles across at this point. And this is tiny compared to the mouth of the Amazon, which has places you can't see across. It disappears over the horizon. So this is an enormous forest. But the first thing that strikes you when you walk into a forest like this, or at least the first thing that struck me, isn't the scale of it, it's the density of a forest like this, the amount of life that you can pack into a very small area, the amount of life that can exist in a square meter or a square kilometer, is kind of hard to comprehend. And what actually greets you, then is just a wall of green. You walk into a tropical forest, unlike if you walked out here in Cary, where you might see 100 feet in any given direction, or at least 10 or 20. I mean, you're lucky to see 10 feet in a tropical forest.
And it's that density of life that really makes tropical forests super unique. It leads, once you start looking at that forest, to seeing not just the amount of plants that exist there, but the diversity of that life. There are so many species of plants, animals, insects, fungi, bacteria, and it's really what makes these forests unique is their density.
So what makes this forest able to have this giant density, this density of life within it, ultimately, it's its productivity. So when we talk about productivity for tropical forests, what we think about is essentially photosynthesis. In here we have a figure that looks like a heartbeat, and it's often referred to like the heartbeat of Earth. This is capturing essentially photosynthesis. It's called gross primary productivity. It's the amount of energy that our terrestrial ecosystems can pull from light and then carbon out of the air into their tissues. So as we go from red colors in small pixels to green or blue colors in big pixels, we're going from low productivity to high productivity. And this is cycling through the months of a year. It's actually cycling through 10 years in sequence. And what you probably see is where we're from, in these northern hemisphere get really small and really big and really small and really big, but when we look in tropical areas, they go from big to really big and big to really big. And when we add this up over the course of a year, the totals end up with this massive difference where these tropical places have this density of life, because they're so incredibly productive.
And we can put a couple numbers on this, to sort of put this in context, tropical forests only represent about 10% of land area on Earth, but this debate only representing 10% of that. They are 60% of that productivity. We just saw that's your land based carbon uptake. So that's a huge imbalance. I mean, they're really, really important to then carbon cycling and really just global climate functioning on our planet. They also just turn over really quickly. So they're about 40% of biomass carbon storage, but that's still enormously disproportionate to their land area. Now it's not just that tropical forests matter right now, or that they're different from other ecosystems, or that they're more productive, they've also mattered over time. So if you take all the carbon we've emitted through burning fossil fuels, coal, natural gas, all these things into the atmosphere, About 70% of that is still there. It's still in the atmosphere, and about 30% has been consumed by stuff somewhere on our planet. And depending what numbers you look at, it's probably closer to 10% but maybe 10 to 15% of that is inside of tree tissues in tropical forests, and that's CO two that otherwise would be in the atmosphere, affecting our planet, warming it, making conditions even worse for us. Now I'm going to talk a lot about carbon. An ecosystem function, because that's really what my lab focuses on a lot.
But I want to highlight the diversity of these forests is also incredible. So anything we talk about related to this directly ties into diversity as well. Diversity is a lot harder to put exact numbers on, but probably half, maybe a little more, of species that exist on Earth exist in this 10% of land area as well. So they dominate huge aspects of forest function. Now, when my research group goes out into a forest, we sort of fight through all the vines and the bushes and the trees and everything like this, all the mosquitoes and bugs and spiders, looking through all these living trees, trying to find trees that are dying. And what we're really interested in is what controls tree death. Because I mentioned these forests have all these functions, and those functions are tied to the trees, and basically the cessation of those functions. So they capture all this carbon when it's released. It's released because they die, they're supporting all this biodiversity.
They aren't serving that same function when they die. Now, tree mortality, like these trees, right here, is normal. I mean trees die. That is a completely healthy aspect of forest function. Now, what becomes concerning is if tree death, events like this start to become more common, they start changing, because that might change, then how our planet functions, particularly in a place like a tropical forest with these big planetary and unfortunately, we know trees in these tropical forests are dying more now than they did in the past. Does anybody here want to hazard some guests about why tropical trees are dying more now than they used to? There are a lot of right answers. What was it? Water, temperature, temperature, acidity, maybe bigger storms, mining, nobody says mining, that's a great one. Disease, air pollution. I mean, these are all all great answers. They're all very valid things.
You know, these are part of the trying to understand why trees are dying. A big one that comes up front and center, I'm kind of surprised it wasn't mentioned more, although mining did make an appearance, is us. We cut trees down for a lot of different reasons. So one of the main reasons that trees die is we cut them down, we destroy those trees. So deforestation happens for lots of reasons. We might directly just want to cut the trees down, but more often, we're trying to do a different thing, and the trees are in the way. So we have commercial agriculture, livestock, ranching. Mining is a fascinating one. If you watch the news here, I don't think you hear about it as much, but if you go watch the nightly news in Brazil, mining will come up all the time. Mining is the number one illicit activity that now happens in South America. It is an enormous, I think, many, billion dollar industry illegal mining in the Amazon and huge swaths of forests are cut down, and then mercury poisoning happens across big areas. It's a crazy, crazy thing. If we add up the numbers, they're also pretty staggering.
So in 2023 an estimated 9.2 million acres of primary, previously uncut tropical forest was cut down. This works out to around 10 soccer fields of forest per minute all year. It's only 10% of the planet, but it has this huge effect on the planet. We're cutting it down at that really fast rate. Now this is super important, but ecologically, there's not actually that much more we can do here. People have developed the tools where we can map this from space. We know where it's happening, we know when it's happening, we know the consequences. There's a lot to be done to figure out how to restore and things like that. Instead of knowing that deforestation is happening, where, as ecologists, people have done a lot of the work, it has shifted into more of a socio economic political issue, where, like, people need to pull certain levers to change our economic priorities, or whatever has to happen. It's a little bit out of the ecologist hand. But what's concerning for us as ecologists is when we go away from deforestation, we still see trees are dying more so even in areas where there is no deforestation, trees are not being cut down. Humans are not damaging these forests. We see increasing rates of tree mortality over the last several decades. So we're only going to have a couple data slides. This is one of them right here. Our x axis is from around 1985 until 2010
or 2015 actually, it goes a little beyond that axis. And our y axis is biomass mortality. It's it's tree death rates. But with people, or, you know, most animals, they have relatively similar size, you could just put a percent number. But because trees are so enormously different in size, you kind of need to come up with a common currency. So we calculate their biomass to understand their effect on ecosystem function, and add up how much biomass dies per area per year. This is from the Amazon, hundreds of forest plots all across the Amazon and over. The last three decades, there's been this consistent trend of increasing tree mortality. There's a little weird bump here. This is when we didn't have that many plots. So the best data are really like 1990 forward, where the line gets darker. But regardless, we see this trend increasing tree death rates. Similar data have been published for other places, Northern Australia, where we are not cutting trees down, they're still dying more presumably, because our climate, the conditions these trees exist in, is not the same as it was when they first started living there. So this is a big concern I mentioned earlier. These tropical forests have this really important function. They accumulate all this carbon. They play this disproportionate role in planetary carbon cycling, and if they're dying more, that's going to make it more challenging for them to continue storing carbon. And there are a lot of concerns right now that even where we aren't, or even if we stopped deforestation, we would start having these forests emitting carbon to the atmosphere instead of pulling it out of the atmosphere. And this really leads into the driving question of our research group and a driving question of the gigante project, which is when, where and why do trees die? Somewhat remarkably, we don't know this answer as a baseline, much less why those trees are dying more than they were in the past. So we need to know how it operates currently to start understanding how these trends change over time. Oh,
sorry, froze there for a sec. So we can start thinking about places where we do know a lot about changes in tree mortality, and think about some of the usual suspects. So if you live in the US, we can think about two major ones that a lot of you probably think about when you think about trees dying just watching the news or reading papers here. One would be fire, and the other would be pests or pathogen outbreaks. Now these are really important here. Increasing fires are decimating the western US, parts of Canada, and you'll hear about fires in places like Brazil. A fact about them, though, is they're they're really part of deforestation. People cut down patches of forest, they light that material on fire, and it may spread a little bit into the edge of a forest, but you're never having canopy fires. You're not having fires in forest interiors. That trend I showed has no relationship with fire, and we think here a ton about pests and pathogens. And if you go to a tropical forest, you'll be like overwhelmed by the number of bugs and possible pests and things like that. But we don't see outbreaks, and it probably has to do with the diversity. You probably have so many species. And there are some data suggesting this is just true. That's a hard thing to prove that. Whatever the pathogen is can never get to enough of a density to then overwhelm healthy trees and like, wash across the landscape, because there aren't enough trees to eat. So these, it's actually kind of marked. There are no recorded, documented outbreaks in a mainland tropical forest ever. It's something we just, we just, we just don't really see. So ultimately, we do not know why these trees are dying in tropical forests, and we don't know why it's increasing.
So let's address that. Why don't we know it seems like something that we should be concerned about, the people should be working on, and we should, should have some idea. And I and people are working on this, and they're trying to solve it. But the issue is, it's it's hard, it's hard to figure out what's going on. And there are two main reasons. The first is the diversity of these forests. And I want to use a place we work, in Colombia, as an example. So this is a map of Colombia right here, if you don't look at it often. The main thing that's important here is Colombia has a little boot, a little foot that sticks out down here at the bottom. And this green thing is amaki Aku national park that is the Amazon River that runs there. This is there. So Columbia has that port of access to the Amazon. And ama kiowakoo is a super beautiful, amazing place. And in it, it has a 25 hectare forest dynamics plot. We'll talk about them a couple times today. Those are patches of forest where people go measure every single tree, and they do it repeatedly over time, to understand how forests change over time. So 25 hectares is about 62 acres. That plot has a little over 1300 tree species in it. If we compare that to all of the Continental, contiguous United States, which is a little under 2 billion acres. So Earth 38 million times larger than that plot. There are only 880 species. The diversity is really hard to wrap your mind around, and it might be tempting like, All right, well, maybe that just has all the trees in the Amazon, or, I don't know, maybe it does. Well, we can also think about how diversity changes over space in the tropics, and it actually changes faster than here. So this is a place near and dear to my heart. I lived for more than half a seven years on that island in Panama. There are, this is, this is central Panama. It's really like the best studied place in the tropics. There are plots all across this isthmus right here. It's only about. 30 miles from that forest plot down to the forest plots here. Now, as we move from this plot here in San Lorenzo to borrow barrel Colorado right here, those two forests have almost no species in common, 100 species, none shared. We go another 20 miles down near the drier part of the isthmus right here, also no species in common, so you have this remarkable turnover of species as well.
So to understand these forests, understand how they function, you would need a massive research investment to figure out how they work. It doesn't necessarily mean that the investment needs to scale with species diversity, but it would be a little more than you would invest here to try to get to the same same answers. And this gets to our second problem. Research investment in the tropics is comically low. This is our second data thing. There's a lot going on here, so we're going to walk through it slowly. What we have are three types of data, the green represents total forest or total vegetated areas, like forests, Savannah, shrublands, places where you have woody plants. Our blue is the amount of carbon in that area, and our red is the basically amount of effort we've put into studying those places. It's ground inventory density. It's talking about those plots I mentioned, and that's our main tool for studying forests. And our y axis is more of all of those things. And then our x axis is latitude. So basically, this is the end of Patagonia. This comes up to the equator. Panama is like, right there. Brazil is in this general area. And then we have the US, North America, Europe, on this right end. Now start with our green line. So this line tells us, for each measure, each number of latitude, how much vegetation area. So it would be like square kilometers of area on the forest. But all you need to know is, more is higher is more. So you see there's a little bit in the southern hemisphere, and then at 23 degrees latitude, there's this huge spike that is the start of the tropics. And then right around the end, it drops off again. And then we have a lot more vegetation area. This is boreal and temperate forests, the US, Canada, Russia, places like that. Huge forest area as well. Earlier, though, I mentioned this density of life, the amount of material, stuff, living organs, that are packed into a square meter, and it's higher in the tropics. So our blue line has this really high peak near the equator in tropical forests. And even though we have a lot of area in the northern hemisphere, there's actually quite a bit less carbon because it's not as dense in those regions. And now you've probably all gotten to my take home point already. These red or kind of purplish marks here, those are our ground plots how we study these forests. And you'll notice between 23 and 23 there's a pretty big hole in where we're studying tropical forests, where we're investing money in setting up research plots to to investigate them. So these forests matter. They're dying more, and we do not have the infrastructure that we need to study them. So how do we start to solve this problem? Like what's a good starting point when we are so far behind our knowledge of forests like we have around here because of institutes like Cary, we understand the Northeast very well. How do we try to build this type of knowledge for this really important ecosystem? I guess I want to note there are incredible people doing incredible research all across the tropics. It's the amount of resources available to those people that differs between here and there. Now we're starting to try to solve this problem by focusing on giant trees. I'm going to Wait one sec while this very slowly loads.
So the reason we focus on giant trees is they have a giant contribution to their ecosystems. Pun, very intended, right there. The largest 1% of trees in tropical forests typically have about half of the biomass. So biomass is a pretty decent proxy for their contributions to function. And this isn't like a ballpark. We know this mathematically. People have actually done the measurements. They are responsible about half of carbon storage, half of carbon sequestration. They dominate ecosystem function. So we think back on that number I mentioned earlier. We saw that graph, the heartbeat of the planet, all that gross primary productivity. 60% of the productivity on Earth is in tropical forests. Well, that means that these 1% of trees are about 30% of the planet's primary production. So if we can start to understand these trees, we can make major headway in understanding these ecosystems, as I mentioned. Four it's not just carbon. There's a lot more to this tree. They have huge impacts on biodiversity as well. Giant trees are generally regarded as these linchpins of diversity in these ecosystems. It ranges from monkey troops using only specific species, to total fruit crop to their decaying material, their leaves, space for habitat for animals and plants. They shape the soil communities, and they have these enormous effects on the ecosystem. Just one anecdote that I think is one that stands out as a weird but very memorable one. I have a good friend who studies ants, and he studies ants in trees, and across around 250 trees in Panama, his work found more than 200 species of ants that only live in trees. They largely live mostly in the big trees, and they never come down to the ground. And like beetles, would be way more diverse than the ants, like all these things would be packed into these big trees, they shape huge amounts of the life in these forests. So I'm telling you, giant trees are important.
But why? Why are giant trees so important? It really comes down to a main fact, they keep growing. We have determinant growth. It's like the biggest person you know is like three times bigger than you are, maybe four times bigger, not that much bigger. Tropical trees, or trees in general, these large trees continue growing for hundreds or even 1000s of years. And because of that, the largest tree in a given forest can be 1000 times bigger than the average tree. And it's this indeterminate growth, this continued growth over hundreds of 1000s of years, that leads to this really important function.
Now I want to show you a video here of a place we were just in, in Borneo, about two and a half months ago, because this is the tallest tree in the tropics, and I think it captures tree size really well and how big they can get before we watch this video, I want to note as we walked up to this tree, we were being led by Tom buchalos, who's a research assistant in drone pilot we work with. He's so cool. He's not telling us which tree it is, and they're all so huge, we wouldn't know that this is the tallest tree, because there are a ton of trees almost this tall in this forest. But let's watch a video of this,
and I'll go I uh, unfortunately, we have like, 15 gigabytes of videos in this presentation. So this part down here, I'm standing probably 25 feet from the trunk, actually, because I can't get closer, because that's how big the root buttresses are. The Statue of Liberty would come to like, right there. This is 100.8 meters tall, so about 330 feet. And like I said, there are tons like this tree is just about as tall. I couldn't tell you from the ground that one is is taller and the other shorter. So these organisms get this big. So I'm mentioning these huge impacts on function, if we think about that, compared to the tree like see out our back door here, it kind of becomes no wonder that if you have 1% of your trees that are this big, you might have a really big effect on global function. So we see these trees are big, they're important. But why don't we know what kills them? I mentioned this tree mortality. We're really interested in we want to try to answer that. Try to answer that question. Why don't we know the answer? It comes down to the fact that there are too few trees, and they're sampled too infrequently. I mentioned it's 1% of trees. So if we're out there and we're measuring all the trees, how we typically study forests, for very good reasons, trying to understand ecosystems and communities, only 1% of our data are relevant to understanding the fate of these really important organisms, and it's really hard to do typical research on these forests, so you can't get there very frequently and use using traditional methods. And this is where gigante comes in. The goal of this project, fundamentally, is to reveal when, where and why giant tropical trees die. So it ties into that question I mentioned earlier, that is the main drive for our research. This is also a drive for gigante. Now, this project is a big collaboration. I mentioned a major contributor, who's the CO leader of this project. We have a European lead on this too. There's Adriane ESRI al mulber, who's actually from Panama and then Brazil, and now lives in the UK and runs the other half of this project the University of burning Birmingham. Now, Adrian and I have been thinking about this problem for more than half a decade, and we're really excited about gigante now, because we're starting to address these issues. Is, but I want to highlight sort of where we got to and where we started trying to think about and we knew we didn't know what killed giant trees, and we wanted to try to solve this problem. And we started thinking like most forest ecologists do, about plots. So a typical forest plot is one hectare. That's 100 meters by 100 meters, like two and a half football fields, the largest forest plot that exists is 50 hectares. So that's a kilometer by half a kilometer. Now, in a typical plot, fewer than one giant tree dies in a given year. So if you want to understand giant tree death, you need 1000s of deaths in many different conditions, many different times to understand what are the drivers over this if you get fewer than one a year, and you gotta look your 50 Hector plot, which is has been revolutionary for understanding of forests. Has been really, really impactful. You're still getting 45 dead trees of that size, maybe 50. It's not it's still not enough now to set up and run a 50 hectare plot. So what they do with this is they go out into a forest, they map an area, build out all these grids, a kilometer by half a kilometer, and they measure every stem over one centimeter. And it takes a year for 10 people to do that, identify the species, map them. It's incredibly expensive. It takes another year and a half or so to curate all those data and get them ready for analysis. And then you wait two and a half years and you start again. And that's, I mean, it's an incredible, unbelievable effort. We've learned so much from it, but we can't learn what's happening the giant cheese from that. We did some calculations when we were trying to figure out how to use plots to answer this question, and we realized you would need at least a 750 hectare plot to get at an annual mortality rate, because you need to measure it every year. It wouldn't be possible, like you couldn't do this. But with Hagan today, we're taking a different approach, to build a plot of the scale, to ask not just how many trees or what's the mortality rate this year, but how does it vary season by season, or place by place, or driver by driver? And that's why building plots that are 1500 hectares in size. How do we do that? How do you, if you can't do it on the ground, with these these forest plots, how do you build a 1500 hectare plot? What we're doing is using drones. So this photo is actually from last week. Vanessa Rubio, who's a postdoc who's based here at Cary, is in Cameroon right now, setting up the last gigante site with a large group of colleagues and training this group how to operate a drone like this one. Now this drone looks like a plane, but it's an autonomous drone. It's operated by building flight plans on a laptop, and all you do is babysit its take off and landing, assuming you built your flight plan well enough. They're really incredible, and they are what makes this work possible, this advance in technology, and also we have incredible people who have been managing this project and managing these teams, training people, getting them on board. Cesar Gutierrez is our head technician in Panama, and a drone pilot who did a lot of work building out the protocols how to operate. This Panama was our test bed figuring out what does and doesn't work. We did have to train change drone types at one point because it did not work. This is Vanessa, who I just mentioned. This is Ian McGregor, who's here in the audience. Ian built the analytical pipeline for going through all of the drone data. Not pictured here, but Gisele bimuri is a postdoc in Brazil who's doing similar work to what Vanessa is doing, just based in Brazil. Now I want to highlight this drone taking off, because I think that will will capture how it how it functions. Here you can see right now the motors are facing upwards. They're spinning pretty slowly. Vanessa press go, and now it takes off like a helicopter, but it looks like a plane, right? It's spinning. It has a wind sensor, so it's measuring wind direction and speed to make sure it's safe to take off. If it wasn't, it would land itself. It decided it's okay. It's turning. And now it rotate its motors down, and now it flies like a plane. And when it switches that mode and accelerate there, it actually uses way, way less energy to fly. So that drone could fly for an hour and a half and map this forest. And what we do is we just take high resolution photos, like a Sony camera, one you could buy now. It's rigged to a bunch of GPS and things like this, but it's taking normal photos like you might take with a camera at home, and it's taking them in incredibly high density over very large areas, and in about two days, if the weather cooperates, we can map an area like this that's 15 square kilometers, and we can map it depending on our site and some of the laws there, we map them either at two or four centimeters resolution. So that means each pixel is like this big, or like that big for 15 square kilometers. This is a down sampled image. The full one is 50 gigabytes, so you like, could not open this on my computer. Now we do this every month. So every month, a pilot in each site goes and flies the drone and. And collects these data. And there are two things that really come out of it. One is this image. It's all the photos stitched together into a super high resolution image of the area. And we can look at a change from July to August and see a crown defoliating. And this is what Ian has worked on, building an algorithm that can basically tell us, yep, this is the crown without us having to look through the imagery ourselves and this, or we'll talk in a second, our structural gaps are then fed into an iPad that the team can walk into the field and collect the data, right? This is what's happening at this site, and we'll see a video on that in a moment. I think the really cool part about this is the structural part. So we're just taking photos. But if you take lots of photos in really high density. And you could do this at home, if you just took photos with your camera. And you know something about the position where those photos are taken from, you can actually reconstruct the three dimensional surface of whatever you're photographing. So what we can do is we can build a three dimensional model of the entire 15 square kilometer forest, and then we can do it next month, and then we can do it the next month, and we compare that over months. Over months, and when a tree falls, we can measure where that tree fell, and that same information is fed into the field team that goes out and tries to figure out what happened. Why is that tree dying? And this is the part that I get really excited about, is the forensic ecology. We basically go perform these tree autopsies every month try to figure out, why are these trees dying. Now we're going to watch a video here that was put together by a really great team of journalists who came with us to Brazil to see us do some training. I mentioned Gisele before. This is Gisele right here. It's going to be mostly in Portuguese and Spanish, but with subtitles, so let's let you know that ahead of time, but you'll see the general idea, like how we get there, look at these sites and interact with it, and the Portugal Spanish is very we have a very trilingual team. So the conversation mice Portuguese is terrible. So I try to guess what people are saying in Portuguese, but in general, it's a very trilingual group. As we go and work you working
Evan Gora
So we basically go to each of these sites and record what's happened within the last four to six weeks and try to figure out what could have been the cause. Not pictured there, but we record this huge suite of also risk factors, things like, are these trees rotting? Do they have lianas? Many other factors that we know also connect to lots of other research that could help us understand multi factor events that lead to tree mortality, because that probably is the reality in many times, not just one thing, it's several things adding up. I mentioned before, our forests also change a lot over space. They differ from place to place, so we can't just work in one place, and hope to understand how tropical forests function or how trees die, we need to capture some variation. So we're deploying this in five different sites across the tropics. The place where we started mentioned before was borough, Colorado, Island. This is the best study place in the tropics. So if you're going to test methods somewhere, it is the place to go. So that's where we started. It also has a very moderate climate among these different sites. It has a dry season, but it's not that dry. It's warm, but it's not that warm, it's not cold. It is interesting. It has a lot of storms. We have this central Amazonian site that's very wet. It has more than 10 centimeters of rain, and every month, every year, it's like very wet all the time, some more than others, but it's a very wet Amazonian forest in Fauci Vera Cruz, this is the driest, hottest forest that exists on earth. So if we want to capture the extreme of climate, that'll give us some idea how that type of tree dies or survives that type of climate. In Cameroon, in Malaysia, we have completely different. Evolutionary trajectories. So those forests diverge from American Forests a long time ago. Here you have really large numbers of large diameter trees, but they're not as tall as the trees you would have in danden Valley, where you have the tallest trees in the world, or at least in the tropics, but there aren't as many other individuals. You have some interesting climate things there. So try to capture really broad variation in how these forests function. Now, as we're doing this work, we're also toward the future of this gigante project, and we have some cool things that have been arising recently. We've been working with a couple different groups trying to figure out how to leverage artificial intelligence for this work. So just on Tuesday, we were sent data by a group where we can now use artificial intelligence to go through one of these images and outline every single crown for us, which could then make it process. Possible to do individualized time series monitoring with AI. And we can say, You know what? Actually, we just wanted to study the giant trees. And with a click of a button, you can just have the giant trees. This is something we're really excited about. It's just starting now. We also work with a lot of people, and the outpouring from the space born sensor community has been, like a little overwhelming. They're desperate for these data, data on what's happening at scale and in high time frequency just doesn't really exist in tropical forests. So we've been contacted by a ton of people who work with a ton of different missions, asking for aspects of these data so they can build models, test their models. This is just, I think NASA produce is showing all the missions they have in existence and the things that are going to be put out to do monitoring of Earth's surface. We're obviously not working with all of these, but we've had people working on several people working on several of these different missions reach out with interest in using these. So I've talked a bunch about the project. Let's go back to the main question, what's killing these tropical trees, right? That's what we're most interested in. It's what we're trying to answer. Now, I'll tell you, we don't know, because we're just starting still, but we're very confident gigante can deliver these answers, and our intention is to run this project for at least 10 years and measure explicitly what exactly is happening in these forests and how it's changing. But in the meantime, we have been working with these groups that we've been developing this large collaborative network to try to understand what are the main candidates that can be driving this and one of them that many people are working around, many people are interested in, particularly in the physiology world, is drought. Drought, both in terms of intermittent rainfall, but also when it gets hotter in the atmosphere, it acts like a drought for the trees. It pulls water out of them more quickly, and this last year became sort of catastrophic in the Amazon, the temperatures got so high and the rainfall stopped for so long, the places that had never dried up in generations that, like villages, can't remember ever not having a river there, and they always have, and it's their main way to get in and out of their village. It dried up and they could not get in and out a like once in a millennium type drought over the last two years in the Amazon. So this is important, and it's the type of thing we're going to test with our data. But something's been interesting as we've been assembling information about what might be driving these patterns. We've come across a lot of data related to storm cause mortality. When I say storms, I'm not talking about cyclones or hurricanes. Those actually mostly form in the tropics and hit not tropical places in the tropics. Instead, what you have are these everyday, typical storms that happen constantly in Panama, from May to December, at like noon to 3pm there is a storm every day. It is just something that happens. And we know those storms cumulatively kill hundreds of millions, likely several billion trees in tropical forests every year. And we also know that storm frequency in tropical forests has been increasing. The estimates are very variable, but they all show the same positive trend, either five to 25% per decade of increase in storm frequency. That is a remarkably fast increase. And if they already kill a lot of trees, and they're increasing, they presumably they're killing even more. Now this is something that is sort of smoking gun, and we need data to show whether or not this is happening, and we have already encountered this is a tree that was uprooted in Panama. These are trees in Panama killed by a single lightning strike. We're already seeing data that storms are killing a lot of trees in our hegante sites. Now this is data from our stormiest place. So this is, like, very, very preliminary. It doesn't mean that storms are the direct cause of these things, but, you know, we have an idea that storms might be important, and then the first data we have coming in are indicating a lot of storm effects. So we're excited about that. We're excited to keep operating this project and start answering these questions. You know, what is killing these trees and controlling their function? And before we end, I want to. Highlight that we have cool technology, and the technology, I think, often can steal the show, but it is people who make this research possible. We have 13 people on the ground across these places who work every day and make all of this work possible. And it is them who really are the key part of the project. Have enormous expertise in this project, but also in these ecosystems, these teams are really incredible, and keeping these teams functioning, we believe, is really key to understanding the future of our planet. This type of work has to get done by someone or still not understand tropical forests, and we think we're really well positioned to be the people doing that work moving forward with that, I'd be really happy to take any questions.
Joshua Ginsberg
So why don't we start off with a question from the virtual audience. They're curious if the growth rate of the tropical trees are also changing along with the death rates. Yes,
Evan Gora
yeah. So there definitely have been effects on growth as well in tropical forests, and for that, drought and water stress are super important. When a tree doesn't have enough water, it has to close all the little holes in its leaves to keep that water inside. And if those holes are closed, they cannot pull carbon out of the atmosphere. They can't grow. So there are big growth effects. Partisan I didn't mention it is the growth component is much, much easier to study, because you can study every single tree and a lot of its physiological processes. So we've learned so much more about growth, certainly much more work to do. But for mortality, we're in the stone age compared to growth, where we're essentially in space already.
Joshua Ginsberg
Question for the audience, one there, how much do you know about species lifespan in general? How much do you know about species lifespan in general? I'm going to repeat it so people out there can hear you. I
Evan Gora
didn't plant that question, that's incredible. So Vanessa, one of her post doc projects. This is Vanessa right here is asking that question. So we don't actually know that much about species lifespan in the tropics, there have recently been a couple efforts to figure out certain species that make tree rings, but we rely on tree rings for age, and most tropical trees don't make consistent seasonal rings, so you can't use that tool. Now, there have been some efforts to do carbon dating, which has some inaccuracies plus or minus 100 years. But in doing that in some relatively small sample sizes, they'll pull out a tree that's like 1500 years old, and it's like we looked at 50 and one of them was 1500 years old. So we don't know how old they can get, and we don't know how that varies in a site. So the work I was mentioning that Vanessa is leading, we assembled this data set in Panama. I mentioned it has the most data on tropical forests anywhere. We pulled together all the plots that can exist in the area into one big data set. I think it has 9.7 million years of tree measurements in one community. And from that, for the first time, you can really estimate differences in mortality among the big trees, and they vary from an average residence time once they get big. So it might be 200 years till you get there an average residence time of 400 years to an average residence time of 20 among species, and we haven't tested this anywhere else. We don't know. But I mean that variation is staggering for the trees that, on average, live 400 some of those individuals are 1000 2000 years old. They must be so huge variation.
Joshua Ginsberg
Another question from the virtual audience was, do the trees all occur in one genera? Or are there lots of different kinds of giant trees? Yeah,
Evan Gora
tons of kinds of giant trees, a lot of them out there. Now, part of the reason we also pick the giant trees I mentioned, you know, it's a smaller number to work on. There are two other advantages. They are not as diverse as the rest of the forests. So it'll probably be 10% of the species represent the majority of your tropical forest canopy. So it saves us a lot of that nightmare of sorting through 1300 tree species. It's only 130 but it is still a big advantage and and that really makes makes the work a lot easier, all
Joshua Ginsberg
So are there particular giant tree species that are doing better than others?
Evan Gora
So that is an answer we hope to deliver. I would say that has to be the case. There have to be some that are better suited than others. So we do a bunch of work also on lightning. We are we have a paper that will come out very shortly, that actually shows there are certain species of trees that are better off being struck by lightning than not. When they're struck by lightning from one of these storms, it kills their competitors. It allows them to. Grow More it kills the lianas that parasite parasitize them. Well, storms are coming more and more common. If that tree can also survive wind, which almost certainly it can, then that would suggest it would be better off of having with having more storms. So what we ultimately want to do is not just understand what's killing trees. We want to understand how they survive, and then identify the trees that are suited to future climate and prioritize them for conservation, restoration. Currently, if we're doing a restoration project, we don't know which trees to pick to become the big, healthy tree, which is what everyone's aiming for. But we're, you know, we're driving with the lights off. We're shooting in the dark. We don't know what the target is, because we don't know which species would do well in which environment
Joshua Ginsberg
question here? So do you look at root architecture of the trees as well as size?
Evan Gora
We don't, because the data when the data just don't exist, but it would be so cool. There's a hard thing. You can use things like ground penetrating radar to map roots in trees, it does not work where soils have a lot of clay, which is, like all of the tropics, not actually all, but it's a huge amount. So we would potentially need some some methods development to make that that possible. Imaging. Yeah, so all of our flight areas have LIDAR before the drone starts flying. Area giant
Joshua Ginsberg
trees. So when a tree dies in the forest, do you look around that tree for the other trees, and see what the impact of that death would be?
Evan Gora
Yeah, absolutely. So when we get to those sites, we actually record a huge amount of variables about the area. There are two levels of data. One is the site, it's like what happened at that site, and the other is tree level information for all trees over a certain size. So it's not just one tree, it's a whole suite of them. Now it is enormously time consuming to get all the way down to the one centimeter trees. So currently we cut it at 30 centimeters. If we had the time and the money and the personnel we would love to do every tree, it just becomes limiting at some point.
Joshua Ginsberg
So one from from the virtual audience, which I think is more technical, is what kind of computer do you use to analyze these huge data sets that you get when you fly a drone, sure.
Evan Gora
So we have a couple workstations, ones in Panama, ones here, some of the some of the data we can just work through with, like a classic, like gaming laptop a teenager would be really excited to buy. So we have one of those, and all the sites to do initial processing, and then more complicated stuff can be handled on these workstations, like the one we have here at Cary has like, 500 gigabytes of RAM and all this stuff that it can work through a lot of data, and storing that data is the challenge.
Joshua Ginsberg
Are you thinking about picking up field sites?
Evan Gora
Yeah, so we want there to be more field sites. It's something we're already working with certain colleagues to build. So there's a precedent for building out methods for something like this that is needed and then building the methods so people can replicate that, and that's what we're doing right now. So we're going to publish a paper shortly that will have all the code for the entire pipeline open access that people can operate. It'll have manuals for how to do everything in four different five different languages, that someone could then take and build a similar site and join the network. That's happened with a lot of different approaches. So we're hopeful. We already have a few people who are interested in applying for funds to do something similar. We'll see how successful things are. So I'm
Joshua Ginsberg
going to do one more from here, and one more for the audience. You get the last question, but let me do one from here first, are you studying the relationships with various fungi, and some of the fungi enhance or deter the health of the tree. Yeah.
Evan Gora
So the fungal component is really interesting. It's very complicated, if you try to get into species of fungi, but there is an aspect that we think is very interesting that's kind of been a debate in forest ecology forever, is whether or not trees having rotten insides is bad for them or not. And it's really hard to test that, actually, but modern technology is making a little easier. We have a couple different types of sensors we can use to go across living, healthy trees that look intact and see if they're rotten inside and then for all the trees that are dying, we record basically if they are rotten. Rotten or not, and from that, you could statistically estimate the effect of being rotten on your probability of death. So we're really excited to be able to answer that question, which hasn't been possible
Joshua Ginsberg
in the past. And then there's a whole other study being done on the on the fungi, fungi in the roots. Yes, right, yeah. You know, getting below
Evan Gora
ground immediately is like, now we need another $50 million or something ridiculous, like, it's just so
Joshua Ginsberg
a small ask, yeah, last question for the evening, giant
Unknown Speaker
trees and
Speaker 1
giant trees, why are they not
Evan Gora
studied? It's a great question. So there are giant trees all over the world, and our project is specific to giant tropical trees, because people are studying all the other ones. So basically, we're when we saw that figure about where our research effort is and where our forests are and where the carbon is on the planet. We're really focused on filling this really big hole in that we aren't doing the research there. That being said, we are working with colleagues. We're having conversation with a colleague at Cornell just recently about working with them to set up similar methods in temperate forests, because they are new methods and they haven't been employed elsewhere.
Joshua Ginsberg
We could go on all night, but it's eight o'clock, and I want to let people go to dinner. So can we please thank Evan again for a wonderful talk.
All right, and I warned you. You can't say you weren't warned. Evan did not talk about the cost of doing this. Almost all of Evan's research money comes from the federal government, apropos of our earlier conversation, but if you are interested in supporting us, please go to the website, write me personally, grab me in the hallway. If you want to particularly support the gigante project, you can just go into the notes section on the web and just say we'd like to help Evan out. But fundamentally these days, I would urge you to support any of the nonprofits that do work around the world to make the world a better place for people, plants and animals, because it's a rough time for all of us. So on that wonderfully cheerful note, I will say, have a great weekend and come back soon. Applause.
