Spying on fear in the wild: Q&A with ecologist Meredith Palmer

In the late 1990s, a Yale ecologist named Oswald Schmitz and his students made an interesting discovery by gluing spiders’ mouths shut. Schmitz’s team placed the glued spiders in enclosed fields alongside grasshoppers, and observed how both behaved over time. What they found was even when the spiders were rendered toothless, their presence still changed their prey’s actions: the grasshoppers grazed mostly when the spiders slept, they ate less, and as a result they mated less, resulting in fewer offspring. Fear of the spiders was so powerful that it actually changed the ecosystem that the insects lived in.

Would the same be true of lions and gazelles?

“Of course, I can’t go to the Serengeti with a bucket of glue and glue a lion’s mouth closed,” says Meredith Palmer, an ecologist at Princeton University who studies fear ecology — an underrecognized, but significant, facet of animals’ lives.

Working primarily in Africa as well as in North America, Palmer has found a way to scale up Schmitz’s experiment without glue, thanks to some simple, but powerful, new technology.

“Camera traps and other technology have really been the catalyst for this. They’ve allowed us to start asking all kinds of really fascinating questions about species interactions,” she says. “They allow me to study the behavior of hundreds of thousands of individual animals that I’m monitoring across 1,200 square kilometer [460 square miles of] landscapes — and I need very fine-scale, high-resolution information about what these players are doing in this predator-prey game.”

Palmer’s findings are particularly useful to wildlife managers restoring ecosystems where this game has changed, often because large predators have been removed. These managers can now make strategic decisions on how to reintroduce predators based on Palmer’s work.

Getting to these results often requires some creative improvisation, though: from wrangling a camera trap to play a lion’s roar, to dragging cardboard cutouts of cheetahs on wheels behind a truck, to setting up stuffed lions realistic enough to fool the real thing. It might not be what you first picture when you think of the work a scientist does, but that’s OK: Palmer has been working to change people’s perspective of what a “scientist” really is, anyway.

“We’re not all white men running around in lab coats,” she says. “Some of us, like me, are young women with tattoos and purple hair, and I’m still a scientist.”

Palmer spoke with Mongabay recently by phone to talk about predator-prey dynamics, her wish for camera traps that are as well-spec’d as smartphones, and the importance of citizen scientists as conservation faces a “fierce urgency of now.” The interview has been lightly edited for clarity.

Mongabay: Let’s start with your background. Have you always been interested in animals, even as a kid? If so, how did you get interested in wildlife?

Meredith Palmer: I feel like so many ecologists have this “coming to Jesus” moment where they realize their purpose in life is the animals, but for me it’s always been something I wanted to do. Growing up we had encyclopedias and biographies about biologists in faraway, exotic landscapes, studying strange animals whose names I couldn’t pronounce, and from a very early age I decided that’s what I wanted to be when I grew up. I also had family members who traveled a lot, and hearing those stories as a kid was very inspiring for me, knowing there’s so much more out there to figure out and understand.

I started doing my first real science experiences in high school and after high school. I did a lot of citizen science, which is funny because I work very heavily with citizen science as part of my job. I now manage the same citizen science project that was one of my first gateway experiences. Which I think is really cool.

I grew up knowing I wanted to go to those exotic places and study those ecosystems, but what does that mean practically? So, field work was very fundamental for me to figure out what big question I wanted to answer. I worked on projects studying everything from fish to lizards to monkeys to birds, from developmental biology, to evolution, to mating systems and social behavior.

Through all of those different experiences I came to realize that what I was most interested in was species interactions, and how those interactions structured ecosystems. Now I’m interested in what happens to the Jenga tower when you start pulling those species out of ecosystems and how the ecosystem responds.

How did you start studying the ecology of fear?

I didn’t really know about fear ecology early on; I feel very few people do know about that facet of predator-prey interactions. When we talk about predator-prey dynamics most people think of coyotes eating rabbits or spiders eating bugs. We think about the consumptive effects of predation.

That’s originally what I thought I would be studying. When I wanted to go back to grad school, I found that the University of Minnesota had fantastic data sets, and I found the lab that runs the Serengeti Lion Project in Tanzania. They have half a century’s worth of lion behavioral and demographic data, so if you want a data set on predators, this is the place to work.

When I joined, they had just started out a new complementary initiative, which was at the time the world’s largest-scale camera trap monitoring project. We had over 200 camera traps distributed over something like 1,200 square kilometers in the center of this pristine natural wilderness area. By distributing so many of them across the landscape we’re able to get this very fine scale information on the behavior, the activity, and the distribution of everybody, the entire community of prey and predator species.

What do all these observations tell you about this ecosystem?

Fear ecology deals with all the behavioral choices that prey animals make in order to avoid becoming someone’s lunch. As a prey animal you view the world in the form of trade-offs. Your number one goal is to survive and reproduce — we call that fitness. Pass on your genes, have a lot of babies, that’s how you win the game.

As you’re finding food, finding water, finding mates, at the same time you have to avoid running into predators and being eaten. Every decision you make is a calculated trade-off between obtaining food and avoiding predators.

With the information I had from the Serengeti Lion Project I was able to study what we call landscapes of fear. The idea of the landscape of fear is that your landscape is not homogenous: you have water holes, hills, wooded areas, open plains, rivers. And based on how different predators hunt, some areas of that landscape are going to be more dangerous than others. Lions, for example, are ambush hunting predators, so they’re going to be around water holes, where you know that prey is going to have to come visit, hiding behind rocks, in dense vegetation.

As prey, you have mental maps of your landscape, and you have to navigate those trade-offs between how much risk you need to take to get the opportunities you need to survive. I was able to map that out across the Serengeti landscape, looking at this whole community of prey species.

I think of these as networks you need to have a functional ecosystem. One of the fantastic things, and one of the reasons I love working in these communities, is that they’re so complex.

That’s something I find fascinating in your work: the concept that predators can have an effect that cascades all the way down to plant life and soil in the area. Can you explain that relationship?

It’s so cool! People think that predators eat prey, and that’s the end of the story. And all of the classic predator-prey theory is based on, well, a wolf eats 10 deer, and then you have 10 less deer, eating 20 fewer bushes, so you have more bushes.

But — and this is why I think it’s such a relatable field of study, because you and I feel fear in the same way a wildebeest feels fear — I think we can put ourselves in the mind of a wildebeest. Being under constant threat of being eaten has physiological effects and it affects where they go in the landscape.

It may be that they’re not going to an area where lions always hang out, despite the area having the best food. So, they’re not getting the resources they need, they’re not growing as quickly, not reproducing as much, and they’re not eating plants in parts of the landscape. It’s not only affecting their health and fitness, but what parts of the landscape they’re using.

How can conservation then apply these findings to help the wildlife and habitats you’re studying?

I work in a number of reserves now where these natural networks have been degraded. Due to their very large resource requirements and their danger to humans and livestock, we tend to lose large predators before we lose any other species. I study: how does that change the system? I also get to ask a lot of fundamental questions about what happens when we reintroduce predators. Prey in these systems might not have a personal experience with those predators.

In addition, my work in intact systems is trying to figure out what those functional relationships look like. So, do prey animals avoid predators in space — for example, do they consistently never go to dangerous areas — or in time, waiting until all predators are asleep before they go to use dangerous areas? Do they spread out across the landscape, or do they draw lines of places they never go?

The goal is we want to establish self-regulating ecosystems. And that requires having functional ecological relationships between different species successfully restored. If you have a situation where you have, say, wildebeest and lions that have been living together for 50 years but you want to put in wild dogs, wild dogs hunt very differently: lions are solo ambush predators, while wild dogs hunt in packs, and rely on endurance to run their prey down. So maybe if a wildebeest knew how to respond to lions, would they try and use the same tactics to respond to wild dogs? That would be bad, because it would just get you eaten.

I investigate all these questions, and then build predictive frameworks based on studies of intact systems. By looking comparatively across all those systems, you can start teasing apart what character of each predator dictates the kind of different anti-predator response different prey species use. Based on that knowledge, that helps inform decisions of which kinds of predators you can reintroduce to reserves, and things we can do to mitigate what we call “predator pits” — when a predator eats all of the prey and you have to keep putting new prey in.

Reintroducing predators can be controversial for some communities. For example, I know reintroducing wolves has been very controversial in western North America. How do humans factor into this work?

The wolf issue is actually very personal to me, because my North American work has been with wolf recolonization. I wanted to study community-wide, cascading impacts of wolves coming back into a system. We had a site in northern Minnesota that a wolf pack had recently reestablished and where they were breeding. I set up 100 camera traps, and then wolves came into conflict with some of the local communities and they were removed.

It definitely is an issue, and it’s an issue I think a lot about, especially with the work I do in Africa. I think as conservationists, and as conservationists who aren’t local, we can’t come in and say: your system needs a predator in order for it to go back to how it was 50 years ago, because we’ve decided that’s how it should be. Not only because that is morally and ethically wrong and could put people’s lives at risk, but also because that kind of conservation isn’t sustainable. I think you really need to foster senses of stewardship and build local capacity, and understand the system you’re working with. That system has social dimensions.

I work in a place in Mozambique called Gorongosa National Park. Mozambique has only very recently come out of a hugely devastating, multi-decade civil war. And this protected area used to be a gem of Africa, this biodiversity hotspot where movie stars in the ’50s used to go for vacation. During the war it got completely defaunated — essentially anything larger than a mongoose was wiped out.

Some of the most impressive, large-scale conservation efforts in all of Africa have gone into rebuilding Gorongosa park. And I think the really commendable thing is it is done entirely with community support and community buy-in. Just as much money goes into supporting and training and capacity building in local communities, if not more than goes into dumping in 50 new African buffalo. And that’s how this park is becoming so great and successful. I’d have to go on for hours about how successful this conservation has been, but it has been entirely because we haven’t just built up the park; we have also built up the community.

One other method of technology that you’ve been using is the internet, through quite a few citizen science ID projects. Tell me a bit more about those efforts and how they’ve changed your work.

There are so many good things to say here. And I guess I would like to start off by saying a huge thank you to citizen scientists who participate in our projects, because it is so crucial to being able to do this kind of work. To everyone who has decided, instead of watching The Office for the tenth time, that they were instead going to go on our website and look at some pictures of wildebeest — thank you!

We set up the first camera trap grid in Serengeti National Park in 2010, and that was 200 camera traps. Those camera traps run 24/7; we leave them in the field continuously. For the Serengeti project alone, that’s a decade of imagery. We have now expanded this project in five different countries in Africa and in the United States. We have something like several dozen active camera trap grids. These range in size from 30 or 40 cameras, up to, I think our biggest is 400 camera traps.

That’s an unfathomable amount of camera trap pictures. And I can’t plug a camera trap picture into a statistics program and get an answer; we have to turn the images into numbers, and then we crunch the numbers. And the numbers code what animals are in the image, how many are there, and what they are doing.

I once did a calculation that continuously blows my mind. Looking at the data I used for my Ph.D., if I had spent 20 seconds on each picture for eight hours a day, seven days a week, 52 weeks a year, without holidays or sick days, it would have taken me seven years to process one year’s worth of camera trap data. And, in those seven years, more camera trap images would have been accumulated.

Meanwhile, we’re using this data for ongoing conservation. We’re using it to evaluate effects of putting up or taking down fences, or releasing wild dogs, or increasing anti-poaching patrols — we need to know in as real a time as possible how the animal communities are responding to these conservation changes. It’s critically important, as conservationists trying to protect these ecosystems, to turn that data around.

The citizen scientists are absolutely incredible. When we first came up with the idea of, essentially, crowdsourcing data, we were the first camera trapping project on the internet. The participants turned around a year and a half of data in something like three days. I thought the only person who would go on the website was my mom and grandma, because they felt sorry for me, but we now have something like 100,000 volunteers around the world who participate in our projects.

It’s also a really great platform for engaging people in science, teaching people about science, increasing scientific literacy, and giving students an opportunity to engage in authentic scientific research experiences. We try and provide educational material so people can learn about ecology and conservation and Africa and the work we’re doing, which is also a cool way for us to kind of humanize scientists. You know, we’re not all white men running around in lab coats. Some of us, like me, are young women with tattoos and purple hair, and I’m still a scientist. It’s a great way to break down barriers to participation in science, showcase diversity in science, and get people excited about science.

Especially in this day and age, I think the more we can get people to understand science, not be afraid of science, and participate in science…  I think citizen science is such a great tool for making science something for everyone. Which sounds super cheesy, but is one of my favorite parts of the citizen science process.

To participate in our online projects, you don’t need a background in science, you don’t need a degree. You could be a kid, you could be a granddad, you could be anyone. Maybe all of your experience that you have with African animals is you’ve watched The Lion King; that’s OK. We have tutorials, we have filters, we have checks in place. It’s OK to mess up, it’s OK to learn, and it’s very OK to get super involved in it, because we love that. But if you have questions you can connect directly with researchers, which is something I really love. Scientists aren’t on a pedestal, we’re not locked away in a university, we’re people you can talk to, can ask questions. It’s a really great way to connect with people, and I love it for that aspect.

Since those early days of first putting those images up for citizen scientists to process, how have things evolved?

With the help of citizen scientists, we can now process what would take years in a matter of months, which is super. But ideally, because of what we call the “fierce urgency of now” in conservation, we really do need this data turned around really quickly. So, the first couple of years of camera trap data that citizen scientists processed for us, we created a labeled image library, which we could then use to train artificial intelligence on them. We now rely on this system called crowd AI, which is like a cyborg: some of our data is processed by AI, and some tricky data is processed by humans.

We needed hundreds of thousands of labeled images to give to a computer to learn before that computer could then do the task that you wanted to do. Thanks to our volunteers, we had enough labeled data to train these algorithms to help the processing of future data. Which has massively advanced the fields of computer science and AI; our image library is used to train most camera trap image identifiers now. And we’re trying to teach the AI to do some really fancy things.

Your research, obviously, depends heavily on the use of technology. But are there drawbacks to this? Have you found limitations to how far technology can move conservation?

There are several different things to say in this respect. One of them is, with the advent of all this amazing tech, we can get all of this great data. But as we found out with our camera traps, suddenly you’re drowning in data that you can’t necessarily use without having something to process it, or handle it, or store it. For example, with citizen scientists, making sure the data we get back from untrained experts is good enough to use in science and conservation is its own task.

Another issue that I’m heavily involved in is that a lot of technology is not necessarily built for ecologists. We’re not the driving market. There are so many things that I know are possible; like, my cellphone does a billion things I wish my camera traps would do. But my camera traps were built for hunters, so they don’t figure out how fast an animal is walking, how far away it is, or the body temperature of the animal, even though that technology exists. A huge barrier to ecologists or conservationists in implementing technology in our work is that we simply don’t know how. It’s the cookie jar that’s just out of our grasp.

There are now a lot of amazing communities developing online, like Tech4Wildlife and Conservation X labs, specifically to connect ecologists with technologists. There is a huge maker community out there who have the skills and passion, and there’s a pool of ecologists and conservationists that have problems. What we’ve seen by bringing these groups together has been this huge explosion of possibilities and exciting developments, and incredible new tools.

I’m working right now with an incredible group out of Japan to develop an incredible new open-source camera trap that does all of the fun things I want it to do. And I struggled for years. I never would have been able to do any of this by myself. I soldered my first circuit board this year and that was my big moment.

With all of the incredible images that have come out of this work, do you have a favorite?

Oh, yeah. Some camera traps take far better pictures than I ever could, even if I was sitting out there hiding for months. I think I have one: my favorite, favorite, favorite image is this one nocturnal picture of a hippo. It’s a big, massive hippo facing the camera. The hippo turned to face the camera and opened its jaw, so we have a picture of this massive, gaping hippo jaw, right in front of the camera, beautifully lit against the black background, and it is stunning.

I love the beautiful pictures, but we also get some … weird sh*t. We have an image from a couple of years ago — and this is really gross, so apologies in advance. It’s a gazelle who had just given birth, and it’s pulling the placenta out of its own body and eating it. But it’s incredible because I could live in the Serengeti my whole life and never see anything like that.