Global Development Institute podcast
Sustainable Forest Transitions: In conversation with Dr Pooja Choksi
In this episode, PhD researcher Mariana C. Hernandez-Montilla continues a new series of podcasts linked to the GDI's Sustainable Forest Transitions project. Mariana chats to Dr Pooja Choksi, Postdoctoral Associate at the University of Minnesota and co-founder of Project Dhvani, about her work monitoring the impacts of ecological restoration, including the use of passive acoustic monitoring to track vocalizing species in Indian landscapes.
Music I Use: Bensound.com/free-music-for-videos
License code: QVPSSG18TYB4DVRS
Find out more about the Global Development Institute:
Intro music Anna Banana by Eaters
In this episode, PhD researcher Mariana C. Hernandez-Montilla continues a new series of podcasts linked to the GDI's Sustainable Forest Transitions project. Mariana chats to Dr Pooja Choksi, Postdoctoral Associate at the University of Minnesota and co-founder of Project Dhvani, about her work monitoring the impacts of ecological restoration, including the use of passive acoustic monitoring to track vocalizing species in Indian landscapes.
Music I Use: Bensound.com/free-music-for-videos
License code: QVPSSG18TYB4DVRS
Find out more about the Global Development Institute:
Intro music Anna Banana by Eaters
Speaker 1 [00:00:02] Welcome to the Global Development Institute podcast, based at the University of Manchester. We're Europe's largest research and teaching institutes addressing poverty and inequality. Each episode will bring you the latest thinking, insights and debates in development studies.
Speaker 2 [00:00:29] Hi. I'm Mariana. Welcome to the Sustainable Forest Transition Podcast. We're based at the University of Manchester's Global Development Institute, where our team combines socioeconomic and environmental data to address crucial questions about the impact of reforestation drivers on forests and communities. In this episode, we have a special guest with us today. Pooja Choksi, a postdoctoral researcher at the University of Minnesota specializing in understanding how restoration can be used for ecological and social good. Pooja will discuss novel techniques for monitoring the impacts of ecological restoration, including her work with project Dhvani, which uses passive acoustic monitoring to track vocalizing species in Indian landscapes. Pooja will tell us about using tools to evaluate multidimensional outcomes of restoration. Finally, she'll discuss her research on what make restoration projects effective and sharing insights from her research on biodiversity recovery under different restoration methods. Join us for this insightful discussion. Hey, Pooja. Welcome to the SFP podcast. We're very excited to have you here. Can you start by telling us more about you and giving us an overview of your fascinating work studying restoration?
Speaker 3 [00:02:06] Hi Mariana, thanks for having me here and I'm excited to be talking to you. I study the social and ecological dimensions of land restoration. So when I say land, I think of, forest land. On forest land. I use novel and traditional methods to study these impacts. So some of the novel methods, like you mentioned, I use acoustic satellite imagery. I'm excited by technology. And the traditional methods, things like, you know, using doing vegetation plots, and using household surveys to understand people's perceptions about restoration.
Speaker 2 [00:02:50] Oh, I'm very intrigued by the techniques you use to track biodiversity and how you're using acoustics to monitor restoration, but in plain English. Can you please tell us about the acoustic technique a little bit more? What is it? How it works?
Speaker 3 [00:03:05] So essentially we use acoustics to understand all vocalizing diversity. And it's really important to note here that not all species will vocalize. And when I say vocalize, I mean just make any sound at all, whether it is audible or inaudible to humans. We have a very small range, you know, of frequencies that we can actually hear. But their species that vocalize at all frequencies from very high, you know, the bats are really high to, like, really low frequencies that some mammals use. So what we're trying to do with acoustics is open up a different dimension to understand species presence and species behavior. So presence is just, you know, do they show up on our when we record this data and then we analyze these frequencies, and you can analyze a whole range of them really from zero to like whatever 24,000 or it's 48,000 notes, however much you want to analyze. You can essentially look at whether a species is present or not in that data. That's one way to look at it. Another kind of research that's happening is on species behavior. So, you know, it's, that changes in a species vocalization associated with certain changes in habitat, things like that. So what I would say is, is, is that acoustics essentially opens up a whole new dimension, for us to, to think about how species behave.
Speaker 2 [00:04:36] Yeah. This is so interesting. What led you to do this kind of creative research?
Speaker 3 [00:04:42] To be honest, I'm not. For example, I'm not an ornithologist. I'm not an acoustics engineer. I didn't actually do anything related to acoustics before 2017, so it's been a few years. But the reason I got into this kind of work is because I I was interested in restoration, and as you know, when the UN sort of the decade of restoration and we have very ambitious targets, you know, for example, the bond challenge aims to restore 350 million hectares of land globally. So when you're working at that scale, spatially and temporally, because we want to meet these goals by 2030, it's extremely important to understand what the impact of all of this restoration is going to be. And we need to do that at scale, because if we don't have scale, like we'd be using so much human manpower to do all of this, monitoring that we just don't have the resources. So I got into acoustics because it would allow me to study biodiversity at scale, both spatially and temporally. And I thought that was a really important question, in this decade, especially because initially when we started thinking about restoration, people were more focused on the climate angle that we need restoration for climate mitigation. But now, as we're seeing, people are realizing that, no, we need to think about other dimensions which are social, the social dimension and even biodiversity. So, that's that's pretty much how I came to do this work.
Speaker 2 [00:06:16] Well, can you tell us some real examples where you use acoustics to monitor outcomes in restoration projects?
Speaker 3 [00:06:24] Yeah, sure. So, on my PhD research, which I did from 2017 onwards to 2022, I used acoustics for that. So for my PhD, I studied the impact of this one restoration method in central India. So, the habitat, it's a dry forest. And it is highly invaded by this one shrub called lantana camara, which actually comes from Central America. And the British, around the 1800s, sort of brought the plant over as an ornamental plant. I believe the invasion only got really bad in the last few decades. So there wasn't there were known records of it, sort of, in, in forests in India and, and even open lands. It's not just forests, it's also the non forest ecosystems that are being affected. But my understanding is that the invasion got acute in the last few decades to the point where, I mean, so many of my forest sites, you can't even walk through them. You've got to go commando style under the bushes to, to get to places. So it's highly invasive. And what, what the project was the restoration project that I studied was that there was an NGO that worked there. It's got a foundation for ecological security. It had teamed up with the local forest authority and they said, you know what? We need to get rid of this lantana. People want to get rid of it also, it's it's a real hindrance to people's livelihoods. Because there are a lot of forest dependent people there. And so the, the local communities in each of the villages, the forest department and office came together to remove this lantana. They do it in a way that they remove it for three years. And then and then that's when I came in after they'd done the restoration work, and two years had passed. I went in to understand what happens when you restore it, but restore a forest, or rather when you just remove that lantana. I would say restoration here is a value judgment that it's better. So I'm just going to say what happens when you remove the lantana. And I compared it to sites where there was lantana, heavy lantana invasion, and then comparable control sites where no invasion had happened at all. And so that was one that was one kind of that was my own PhD research. But then I also collaborated with folks, mainly in India, in different projects to use acoustics to understand the impact of restoration work.
Speaker 2 [00:09:08] Wow. And by using this technique, what key patterns did you find?
Speaker 3 [00:09:13] So you find some really interesting stuff. I'm going to start with my own PhD research, for example. We found that first of all, the the impact when we used acoustics is kind of varied across different frequencies. So different species occupy different frequencies. Right. And what we found is that in the low to mid frequency range, which is dominated by certain audible insects and birds, we found that, you know, restoration had a slightly positive effect on, on the acoustic space use, which is essentially looking at. In simple terms, acoustic space use would be just how much acoustic activity is going on over a 24 hour period. So more acoustic activity is, we assume, a good thing. That's what the theories, you know, in an acoustic ecology state, we want more, vocalizations there. So what we found is that, there was a slightly positive, higher acoustic space use, but restoration was not significantly affecting, acoustic space use when we controlled for a lot of things. And similarly, in the higher frequencies, we found that, in fact, it was the unrestored sites that were positively, associated with, higher acoustic spaces. So, it was interesting. It was things I wasn't expecting to see. But I would, I would say about my own findings, that it was very soon after the restoration had actually taken place. And I think that these things probably take time. And it's important to follow it's important to track the same sites over a larger period of time. And so I've been trying to go back, trying to go back to those sites and constantly collecting data to see, what will what's going to happen. But, yeah, it was interesting to see that, you know, in the lower frequencies, restoration had a more positive outcome in the higher frequencies. It was the unrestored sites that had a positive association.
Speaker 2 [00:11:31] Right. So fascinated to hear these. So now let's talk more about the logistics. Tell us about your experience in the field.
Speaker 3 [00:11:41] So I would say in terms of, using acoustics in the field, it's kind of interesting because I have from, from the beginning of my, from the, from the time I started using acoustics to now I've had so many funny stories. But maybe I'll tell you one, after I give you a quick understanding of just how this works so you can. There's a wide range of equipment that you can use. For example, you know that different companies like Cornell Lab of Ornithology, and they, they have an acoustics division, they sort of produce equipment. There are lots of there's wildlife acoustics there. So audio mod, I personally try to use something that's small and light because I've got to cover longer distances. And so what you do is, depending on your sampling design, I mean, that's different. I sort of had a design where in, I had about five points at every at a distance of about 500m each. To ensure that there was no overlap in the acoustic data that these recorders were, capturing. So what I did is I, I got to my point, you know, using my GPS, and I would tie this recorder to the tree, and you tie it about, you know, a little over a meter. So, depending on the kind of recorder you have, you could have a multi-directional mic. So you could have several mics, or you could have just one that faces downwards. Often that's the model that that's most commonly available. But essentially you tie the recorder to the tree, and you leave it out for the number of days that you need. You can even change the way that you are recording so you don't have to record continuously. You could record one minute for every five minutes. It really depends on your needs and the species you are going after. What your research question is. But I do have some really funny experiences from the field. So, for example, in the early days, I don't know why. You know, it was a hot summer day in central India. And when I say hot, I mean it goes to 45 degrees in this landscape. But, you know, you're a first year PhD student. You want to do it on, you want to do everything you can, and you're just like, not a day must be wasted. And it was a really hot day. And I got out and I remember that, I was talking to a colleague who was joining me in the field for a few days. She's she's the co-founder of Project One. And we were all scientists together. An AP PhD program. So Sarika is her name. She and I were walking, and I remember, I climbed up the tree like that, climbed up, but just sort of went up a branch, and I put it and tied it up. I put, you know, I put the recorder in a Ziploc bag. I tied it to the tree. Everything looked great. I took my notes off like day of recording did it. I had I had my field book. And I forgot to press record. Oh, yeah. On a hot summer day. So 15 days pass, which is the duration that I wanted to record. And I go back to the recorder and I am like, oh, wait a minute. I didn't even press record on this, this recorder. And I have to go back another day and come back in this heat to go, come get my recorder again. So I would say, anyone out there using acoustics, don't forget to hit the record button. When you when you leave your video at the point, another another time, I had a squirrel eat the rope that was that held the recorder onto the tree. And every time that kind of happens, you, you know, that data is kind of compromised. I just start all over again. And maybe I didn't have to, but again, you know, in the early days, I was so worried about data contamination and all of that. So I would just start all over with that point. But, yeah, it just ate my rope through and through, and I go there and I see my record on the floor, on the forest floor, on that leaf litter. And I'm like, oh, God, I got to do all this again.
Speaker 2 [00:16:04] Wow. These are very wise advices.
Speaker 3 [00:16:07] You know.
Speaker 2 [00:16:08] So what do advantages have you found compared to traditional surveying? Are there any limitations to consider?
Speaker 3 [00:16:17] So I would say there are pros and cons, like there are pros and cons, of all technology, and all human services as well. So I'll start with the pros. The biggest one is that you get temporal and spatial scale, which you would not get in, a traditional biodiversity survey. And so that I thought was one of the biggest reasons I actually got into this work, is because, it's humanly impossible to do so many surveys, and get research done, you know, get an get the data analysis done and everything. So that's why I got into this work. And I would see as time is passing, this technology is getting relatively less expensive. You know, so there's more and more innovation in this field. Recorders have gotten smaller, lighter, more efficient, easier to use, things like that. Now, you know, I think as academics, we tend to use mostly open source programing software and things like that. So there are lots of that now over time again, that have there's been so much development in terms of R packages. The same with traditional ecology. Like, you know, people created packages to help researchers like us get our work done faster. But now that, you know, going beyond R and Python, there are, other platforms like, IBM on where you can upload your data and analyze it. There are different features. And they're trying to work on some AI stuff. Which, I would say I and a few other researchers have been kind of doing behind the scenes. But they've got some stuff up and ready. It's pretty cool. So I'd encourage you to check out all of these free resources. Now, getting to the cons of of this, of this tool. I think the cons the first is that this this is, this kind of work is if you're not a naturalist, if you're not an expert at insect and bird calls, forget that, cause you can't even hear them, right? Forget the other stuff. That's even more difficult. And the people trying to build libraries of these calls. If you're not one of those experts, then this data is just species agnostic to you. You're just if you're not listening to it and you don't know, like, oh, it's species X, y, z that they're calling to you, it's all just sounds. So, and so I want to use this, not to discourage people, but encourage you to work with, experts, naturalists I worked with to experts in, in species goals. And I'm going to name them a shout out to them. Professor Maria and Siddharth Biniwale, who've also been on my papers. They helped identify each and every bird in, in, in the data set. And birds are easier than, say, the insects and the other taxonomic groups. And so. You need that because if you want to do species level stuff and not just soundscapes as a whole, which is species agnostic, you're going to need that help if you don't have those skills. And, good luck trying to understand all of those, warbler calls. Oh, my God, each warbler has so many different callsthat, I would say I would not have been able to do it without the naturalists. And as, as a, as a, as an ecologist, as someone who tries to, to be more conscious of how I'm, how my ways impact the world. These recorders need batteries and a lot of batteries yet, I mean, you know, these battery, recycling facilities aren't available everywhere, so that meant I had to get my batteries out of my field. That's all that going on my back, out of the field sites. And and to a city where I can get it recycled. Some countries, for example. It's still not, you know, those those recycling systems are not in place yet. Very efficient recycling systems. And so if we're using such tools, we are increasing our footprint, in terms of pollution. So that's something to think about before you get into this.
Speaker 2 [00:20:57] Yeah. A lot of challenges for the future. So yeah, talking about the future. Looking ahead. What do you see as the future potential for scaling of this acoustic monitoring globally, like weather through remote sensing or smartphone apps or other new approaches, just like what you were just telling us.
Speaker 3 [00:21:17] So I think there's a lot of potential for scaling up acoustic monitoring. And I think as, again, there's so much emphasis to get things done really quickly in this decade of restoration. You know, and now we're suddenly seeing a lot of people talking about biodiversity credits. And, and there are different things being spoken about in this in this decade. Now, whether you agree with that or not. I think that's another question. But but there is a need for scalable methodologies is what I take from from all this discussion in this decade. So there is potential to scale it up. But like I mentioned earlier, there are lots of challenges. For example, we haven't reached a point. So. We have in some places. So, you know, you've got bird net, for example. And, and a few other groups sort of working on AI for it to understand species goals. And as difficult as they are in some places, people have unlocked it and they've managed to get accuracy up to, you know, really high numbers, on on species goals. And that's great. Because then you, you know, you can really monitor at these birds. It's kind of easier. But I would say that, for noisier environments, like in the tropics, it's really difficult to get good accuracy. And for each and every sea bird species that you're interested in. So that's going to need a lot of heads coming together to find a solution. But, people are working on it. I'm personally not a computer scientist, so I'm always relying on people with those skills. But, yeah, it's a joint effort. Like I said, you need to bring the naturalists. You've got to get the ecologists. You've got to get the computer scientists together to make this happen. But I will also caution that in our rush to try and scale biodiversity monitoring techniques. And, and as we're trying to scale them, we're trying to flatten complexity. Which complexity? That's really important. We want to flatten everything down to a number. That's what we like to do with people. Right? That's what we like to do is, because we need to make decisions. And it's not easy to make decisions when you have 15 numbers to consider. And we're trying to flatten that complexity into one number. And I would caution, I would caution against that because, that's sort of. Diluting what these frequencies are trying to tell us. And so as we're trying to scale, let's try to not lose the complexity of the data as well. That any ecologist will tell you is telling us very important things, and we need to just figure out a way of considering all these complexities and still making decisions on the goal, policy decisions. And I know it's easy to say and difficult to do. But. But I think it's important. Yeah.
Speaker 2 [00:24:44] Yeah. Thank you for highlighting all this work we have to do. This is all very interesting, but, when we talk about the future, I cannot avoid thinking about the impact in local communities. And wonder from your experience, how do you think we could ensure that these new techniques bring positive outcomes to the livelihoods of local communities that are living nearby these restoration areas?
Speaker 3 [00:25:09] Oh, thank you so much for this question. Because. I grappled with this question ever since I started my research, and I would say this, this of this applies to everything, not even just new technology. Don't you think it applies to a wide range of, of methodologies? But specifically with this, I will I will tell you, I went in very excited that, all the and the people living around my field sites are going to be so excited by my soundscapes. And I made all these dynamic spectrograms to show people, and they engaged with it. So I tried to get people excited by it and like, come, come see me, but like, come with me, not come see me, but come gonna come along with me and we can, put this up and I'll show you how this works and things like that. But I remember just being like, they were people excited by it, but they were like, okay, now I'm going to get to work. This is cool, but okay, you do you and we've got to get to work. And so. I think that as scientists, perhaps, perhaps we're so excited by these things that maybe not everyone is. And in a way, I actually think that that's not too bad a thing, because they didn't think that that was, that was gonna be detrimental to them. If I was a threat and if this technology was a threat, someone would have immediately said, you know, no, we don't want you to do this work. Or, they would have been a little more worried about my presence, I guess. But I had a, decent experience. And I've heard it's not the same for camera traps. People generally don't like having camera traps, even in their community forests and things like that. So I think there is potential, again, to create positive outcomes. And what those positive outcomes could be is, when, if and when we manage to figure out a system, of rapid biodiversity measurements using acoustics, I think people can be trained in these methodologies and then it's in their hands, you know, to to do all this work, maybe just run it through certain soft, like if we can prepare certain code and soft on a software for them, they just need to run it. Analyzes not analyze as numbers but sort of report those numbers. And, and have to cut us out completely. That is if they're interested in, in, in this work and using this method as a, as a measurement tool. So, I think there is potential for positive outcomes, if they're interested in using this technique.
Speaker 2 [00:27:59] Yeah. Thank you. Yeah. So from what I've heard so far, some of your research is trying to help to ensure that habitat restoration is beneficial to people as well. So I also wonder if, have you found some models of restoration that were successful, that were successful and then you think could be applied to other developing country contexts? And if the answer is yes, can you give us an examples? Yeah, sure.
Speaker 3 [00:28:27] I've always been on the research side of things. Wherein I went to these restoration sites and sort of gave an unbiased view of it in academic literature. So from that. So now it's. So, speaking as an outsider, I'm not a practitioner, so maybe my my views are limited. Something I saw in the projects that I studied was that. There was. So, for example, in central India, it was kind of interesting to see different stakeholders come together, to make restoration happen. And I thought it was important to have that open communication, to make it work. I also, you know, there was a, I also thought it was important to consult people about their perception of forest degradation, because I think forest degradation can be defined in so many ways. As ecologists, you know, we might think that something is degraded, but maybe as someone living there, it may not be degraded. It may have value. And so it brings us to the question of what is restoration. And I know that ecologists, again, are grappling with this question. It's a you know, we have dynamic systems. How do you decide these things? But I think it's important to get people's perspective on what they think is degraded and hence what they think would be a good restored ecosystem. So I thought that that's something I saw people do well in some of my sites. Another thing is spillover. You know, spillover effects are so important. And I think that oftentimes our research, doesn't, you know, sometimes, at least in my case, I couldn't account for spillover effects, in neighboring, places. So I would say that's a caveat of my work, for sure. But, for example, when you're trying to restore a space, and this is not this is just a hypothetical example. If you're trying to restore a particular site, and you decide to cut it off and say, like, let's not use this place for a year, what's that going to do to the other sites around it? You know, that use is going to have to come from somewhere else. So in the context of my in the context of the work that I studied, we found that people a villages that was surrounded by these like, it's just don't tell as far as you can see, it's lantana. It's just all barren land, farmland, forest land. Everything was taken over by Lantana. You can. It's just like hills and hills and hills of it. And so people were reliant on it. What do you do then? They needed certain natural resources, and they relied on lantana for their things. And so one of one of my sort of recommendations to, to this group was, that if you're thinking of large scale lantana removal, then we need to think of sustainable alternatives to give people so that we don't have spillover effects. That's something that would need to be planned so that, people need to use things they need as fuel, firewood, things like that. So and if it was coming from the lantana, even though lantana is a really poor source of ignition, it a fire. Sorry, it's a poor fire. A fuel would, we'd have to account for people's needs before we go large scale with restoration projects. Another thing I realized from my own work is that social or sort of human well-being, and ecological outcomes, or what we want to consider benefits may be occurring on very different timelines. And we need to report these differences in timeline. You might see that people get people get some benefits. Say for example, in the available in the sense of the availability of NFB non timber forest products, which could be just fruits, leaves, flowers, things like that. Quickly when restoration happens compared to say, the recovery of biodiversity. However you're measuring it. And so something I felt like a lot of the restoration work, the research that looks at multi dimensional outcomes, I felt like I hadn't read enough about it was, the difference in the timelines on which we see the different outcomes. So we need to be patient. We can't expect everything to happen all at once. And, we love to think of everything as a win win. You know, it could be it could be a win win. But it takes time. Not everything is going to win at the same time. Right? So. Yeah, I would say that's largely that's largely. Things are found. I feel like I didn't directly answer your question about what made them successful, but I think that these were learnings from from the research. And I think these were learnings that I felt like people I presented this work to, the practitioners are very receptive to, to keeping these points in mind, going forward, that could make. That could make it a success. And I think this is applicable to several other middle to low income countries. Even though my experience is largely in India, I can imagine this kind of applying to a lot of other countries as well.
Speaker 2 [00:34:17] Well. Pooja. Thank you for this incredible examples of restoration, empowering people and for showing us tools like the acoustics to paint a fuller picture. I could listen to you all day. But for people working on ecological ecological restoration projects as well as research. Do you have any last thoughts?
Speaker 3 [00:34:40] I'd say perhaps, when we're when we're looking at restoration, let's try to use, an approach that is multi-dimensional in terms of understanding its outcomes. I think often we tend to just, stick to one or the other. Like, you know, where looking at vegetation regeneration and not looking at, you know, like facade regeneration, helping people. Not sure. Of course, this is in the context of spaces where people use natural resources. I guess if it's a protected area and we use is allowed. Of course, the context is completely different. But but I would say those are my last thoughts. As someone who tried to work on different dimensions of it, and I thought it was really. Fascinating and eye opening, and I'm glad I did it and I took the effort to do it. So yeah, I think that's something to keep in mind.
Speaker 2 [00:35:39] Oh. Thank you. I've learned a lot today. Well, now to conclude this fascinating episode. I've been waiting for this question all day. What is your favorite sound of the forest?
Speaker 3 [00:35:52] Oh, I have so many, so so so many. But I will talk about one of them. Okay, so as you know, India is home. Do large cats, tigers, leopards and and where I was working so I wasn't working in a protected area. I was working outside in unprotected forests. And, even there, you know, as I'm sure you've read, they're they're animals. They're roaming about freely. So, I remember this one time, this one recording, actually, and I have many like those, but, I'm just thinking of this one specific one where you you can hear the alarm calls, of this one primate species called the language. So whenever there's a, large cat on, on on the move, these languages start producing alarm calls, because they know that they're probably on the move because they're on the hunt. They go to hunted deer, mostly. And so you have language giving these alarm calls, and you also have certain duas, like the sambar and sambar deer and quail. Deer, calling. And it's so exciting because it's like being on safari, but just acoustically, and, and you're just listening to the drama, you're not actually visualizing it, you're not seeing it in real. So which means you can imagine it. Right? You, and and I think my imagination runs pretty wild. So, I, I love listening to these do my own recordings and creating an image in my mind and creating these, thinking about what what could be happening when these recordings were made. And I and I love those that especially when they're out and, and all the alarm calls are going off. It's so exciting. It's so exciting.
Speaker 2 [00:38:07] Thank you. Pooja. Now, I'm a bit jealous of your work.
Speaker 3 [00:38:12] It's been a.
Speaker 2 [00:38:12] Pleasure having you on the show. All the leaks on the project and papers we had mentioned are available on our website. Thank you for our listeners.
Speaker 3 [00:38:21] Thanks so much, Mariana. It was so nice chatting with you.
Speaker 2 [00:38:24] Thank you. Bye.