Naturalist Selections is an interview series produced by the American Society of Naturalists Graduate Council. We showcase graduate student and postdoc authored work in The American Naturalist, a premier peer-reviewed journal for ecology, evolution, and animal behavior research. Catch up on exciting new papers you may have missed from the journal, and meet some truly brilliant early career naturalists!

In this episode, Femke Batsleer talks with us about her new paper Batsleer et al. 2022: ‘Behavioral Strategies And The Spatial Pattern Formation Of Nesting.’ We chat about digger wasp behavior, building natural history-grounded models using the inverse modeling approach, studying complex and context dependent behaviors in wild populations, and more. You can read Femke’s full paper here: https://www.journals.uchicago.edu/doi/10.1086/717226.

Buzzing with more questions? Email Femke at Femke.Batsleer@ugent.be!

Listen to Naturalist Selections on Spotify!

Credits

Featured Guest: Femke Batsleer, University of Ghent

Host, Editor, Producer: Sarah McPeek, University of Virginia

Original Music: Daniel Nondorf, University of Virginia

Transcript:

Welcome to Naturalist Selections, an interview series featuring graduate student and postdoc-authored work in The American Naturalist, produced by the American Society of Naturalists Graduate Council. Today we hear from graduate student first-author Femke Batsleer about her paper ‘Behavioral Strategies And The Spatial Pattern Formation Of Nesting’, hot of the press of the January 2022 edition of the journal. Femke is a fourth year PhD candidate with Dr. Dries Bonte in the Terrestrial Ecology Unit at Ghent University in Belgium. Femke’s research focuses on the digger wasp Bembix rostrata. During their short summer breeding season, female wasps lay a single egg in a self-dug sand burrow. While most females build only one burrow with one larva, a few may construct up to four active burrows during a breeding season. Females then capture prey meals and carry them back to their burrows to feed their larvae. Sometimes, parasitic flies lay their eggs on the larva’s prey meal. When this happens, the fly larvae steal resources from the developing wasp larva, jeopardizing the wasp larva’s survival. Femke maps clusters of hundreds of wasp burrows stretching across many meters of the Belgian coastal dunes. In her paper, Femke and her coauthors want to understand the ecological drivers of the wasps’ complex clustered nesting patterns. They build an impressive series of natural history-based models to compare to field data from Femke’s digger wasp population. Are wasps clustering their burrows based solely on the availability of suitable habitat? Or are wasps using personal or social information to cluster their nests in groups with other individuals? Well, there’s strong evidence from Femke’s models that all three of these things are going on, but maybe not at the same time. I spoke to Femke to dig a little deeper.

Sarah:  I was, of course really excited about the digger wasps, because I love that system. And I love all the classic behavioral ecology work that’s been done with the digger wasps in Europe. And I’m really curious what it’s like to visit your field sites in Belgium. You know, what do you see and hear and smell when you’re visiting the wasps?

Femke: Yeah, indeed, it’s very special to be in the field when a lot of activity is going on with these digger wasps. But for me, the thing that pops out when I remember my field visits, is the mixed smell of sunscreen and sweat. It doesn’t really have to do anything with digger wasps themselves. But that really gives me flashbacks to the field visits. But yeah, the wasps themselves. It’s really amazing when there’s a lot of activity, especially in the beginning of the season. Then the males are really active and also from literature it’s often called a “sun dance.” And that makes it sound very poetically of course, but it’s these males that are a little bit above the crowd flying around in circles and eights trying to defend their little territory they have, where they really look out for any intruder, other males or even butterflies flying past. They just jump really quickly to them. And then sometimes when the females emerge, you can see that because there are so many males that try to find females, they all come to this one female and then you can see flying balls of digger wasps in the trees, and it’s really amazing to see. I’ve seen it twice or three times.

Sarah: Wow, that sounds incredibly chaotic, but also really beautiful.

Femke: Yeah, it is. It’s also just being there in the field, wow. And then in every corner you can see these digger wasps. They’re quite big, so two or three centimeters. And they just are guarding their nest or flying around coming with prey. And it’s really just standing there and watching them. It really gives this feeling of this little natural wonder you can see.

Sarah: And they create these beautiful patterns of nests all over the landscape, which is what you were interested in in this paper.

Femke: Indeed. Indeed.

Sarah: So I was really interested in the inverse modeling approach that you used in this paper. And I was hoping you could tell me a little bit about what that means, specifically, and how you think the inverse modeling approach helps us understand these complex patterns, like the spatial clustering of your wasp nests.

Femke: Yeah. Okay. So the inverse modeling, very shortly, it’s kind of putting the data on the model and not the model on the data, very shortly. But it also has to do with patterns and mechanisms and how you derive mechanisms and patterns. So in a more ecological, like classical ecological framework, you see a pattern, and then you fit a model, and then try to see what kind of mechanisms or hypotheses of mechanisms you can get out of this. And then if you really want to look into mechanisms themselves, you do experiments, where you hold certain variables under control, and then try to do experiments and see what comes out. But the problem is that now we, more and more, we are looking into these very complex systems where a lot of things are going on at the same time, and we want to look at the effects of mechanisms that work at different spatial scales. And then it’s very hard to tease them apart with modern classical experiments. So I used an individual-based model where you model or simulate different individuals and their behavior, and then see what patterns emerge from there. But the thing is that if you want to use this kind of inverse modeling, you really need a good understanding of your system. And also, research already has to have been done on the more classical parts, because you need some understanding of the mechanisms, and if you miss a certain mechanism that might be very important, then your inverse model doesn’t make any sense. So you need this good natural history understanding. And then if you do that, then you can build this model, see what simulations give, and then see what actual pattern in the fields, if you compare it with that, then you can come up with the idea of what the actual mechanisms are that are going on in your system. So yeah, for complex patterns, it’s very interesting because you can integrate mechanisms that work at different spatial scales, that work or that are studied in otherwise very different fields. For instance, in my system, I combine things from behavioral ecology, but also things from landscape ecology. And with inverse modeling, you can try to put them together and see the how, how the bigger picture is? Or could be.

Sarah: I see. So it’s very driven by the hypotheses that you already have from studying these patterns in the field and then applying them to these statistical models.

Femke: Yeah Yeah.

Sarah: Cool! That leads really well into my next question, which is about the individual based modeling approach itself, which I’m really fascinated by. But I also think it’s really, it can be really complicated to build. Because, of course, natural systems and animals themselves are incredibly complex in the decisions that they make and the kinds of information they’re using, and just what they do when they’re behaving in the wild. So when you’re building these kinds of models, how do you decide what properties to give your wasps? And then how do you decide what properties you’re going to leave out?

Femke: Now it’s a good question, because you really have to think this through before you start working on your model or trying to build a model. The thing is, I think it also was very dynamically how these ideas grew, because I was working on the system already from 2016 for my master thesis. That was mainly the field study I did. So I was really already looking into all the possible mechanisms or explanations you could give there. So yeah, I really had a good understanding of this natural history. But then it was not really a point of what to keep in or let out, but more the question of how to simplify things. For instance, I had the social attraction, so the attraction of an individual wasp to other wasps that are digging or the presence of their nests. And the thing is that in the literature, we have already found evidence of selfish herd mechanisms. So when the density of nests or individuals is higher then the individual chance of a wasp being infected by parasites is lower because you just have this mass of wasps. Because if you’re in a lower density than the parasites can always infect you. And when you’re in a higher density, there are not that many infections, sort of a dilution effect. So we were also looking at this, like how to integrate attraction because of these parasites, or attraction because of other mechanisms. But then we decided to combine these in just attraction and simplify the selfish herd mechanism just into a social attraction. Because it would lead to another level of uncertainties and things to parameterize that we didn’t really know how it really worked in the field. or what response functions we could apply on these kinds of things.

Sarah: Yeah, that makes a lot of sense. Gotta keep things simple.

Femke: And it’s also with other things as well, other parts of the mechanisms how to simplify things and not make your model too complex or try to put every detail that is known from natural history or literature into the model. Just try to come up with a model that can reflect the system in the wild and that we could compare and not make too complex because of all the uncertainties. You can’t fit the data anymore if you just put too many details, then you need millions and millions of rounds of your simulations.

Sarah: Very true, logistical reasons to keep it simple.

Femke: Yeah, indeed. Indeed.

Sarah: So getting a little bit into the actual results that you found, I thought it was really fascinating that your model showed that individual wasps, when they’re selecting nests, are either using that more social attraction information, or they’re using site fidelity information. But you didn’t see any evidence that wasps were using both in making their nesting decisions. So I’m really curious about your thoughts on that? And do you think that this would be an adaptive strategy for the wasps to filter out certain kinds of information over others?

Femke: Yeah, it was also a part of the results that interested me a lot. And also when I was building this model, I found it neat, and also when reading into these behavioral strategies. And I think there are two kinds of explanations. So also I haven’t figured it out myself yet, about how this works, or what effects are, or maybe this is flexible, maybe, in some contexts, like when the populations are less dense than this mutually exclusive behavior isn’t there. But I think, first of all, I was thinking of the physiological constraints. So that’s, you see it a lot in behavioral studies that when a certain trigger is there, then the brain of the animal just follows this trigger or has a reaction on this trigger. And so I think, then it could be possible that it is just physiologically constrained that if you react on a certain trigger, then you react on that trigger, and you can’t combine. Also, for us, I guess, in daily life, if you’re hungry, you’re hungry, and then you act on that. So I think this physiological constraint could explain this, the splitting of the behavior, but also the specialization of strategies is also known in ecology and behavioral ecology in general to exist, and especially when competition is high. So I think in general, it’s good for the fitness or the outcome of fitness, when you have a specialization because you’re competing with other individuals. If you’re specialized in one kind of strategy, then if you would combine them, you would be competing with the different individuals at the same time, it would actually be a lower fitness you would have. So it’s kind of this disruptive selection, I think, that either side has a higher fitness.

Sarah: I think the idea that there are two different strategies the wasps in the population are using is really interesting. How do you think that could relate to risks of parasitism? Because I know you didn’t put that directly in the model. I know, it’s something you’re really interested in.

Femke: Yeah. I think there can be a lot of trade offs going on there. Because if you have this conspecific attraction, then you also lower your chance to get parasitized. But then it really depends, I think, on if this conspecific attraction in relation to parasites is a reaction to the direct presence of these parasites or if this is more like an evolutionarily evolved trait. Then it really depends on the level or like the trigger of the selfish herd mechanism that’s also part of this conspecific attraction, you know?

Sarah: Yeah, I guess one thing to test is whether wasps are selecting nest sites based on prior knowledge of parasitism, or whether they have no idea what’s going on with their offspring, and they’re just picking the best nest sites that they can find.

Femke: Absolutely. When I started working on this model, I was like, now I will understand the mechanisms behind the patterning. It’s always like that, of course. But then so many other questions come out. How does this really work? And you will never know. But that’s, that’s, of course, the very interesting part of doing science. You just come to the next node of understanding a little bit and so many other things radiate. Questions radiate from this. So yeah, that’s it’s really interesting what the real triggers are. What is really built into the behavior of these wasps? And I think, especially because they live quite shortly that I think a lot of behavior is very intrinsic. But then, of course, when you see them being attacked by parasites, they try to make this flight behavior, and it’s called evasive flight performance in the literature. So they do react with just a bit of flying behavior to these parasites being present. But maybe yeah they also then change their behavior of nest site selection. What is the relative importance of this intrinsic behavior and then direct behavioral responses?

Sarah: When you say they’re doing these evasive behaviors, does that mean that wasps might be trying to lead parasites away from their nests by changing their flight patterns?

Femke: Yeah yeah yeah, they’re just flying like crazy around the nest, and then go away again. The parasites are mainly brood parasites, so they try to lay their eggs on the prey that the digger wasp tries to bring to the nest, and then the offspring of the parasite eats away the food so it’s more like a cuckoo style of parasite rather than a parasitoid that directly would kill the offspring of the digger wasp. But you see then if they are trailed by several of these parasites, they just try to shake them off in a way. Shaking off is the normal language you would say for the flight performances.

Sarah: I imagine that’s got to be a lot of fun to watch and track in the field.

Femke: Yeah yeah. Also in general, they’re quite fast. So when they arrive at the nest, they fly slower because they have this prey to carry. Then you can follow them, and then you want to follow them, like, where are they going? But it’s just like, they go up in the air. And then pfft, they’re gone. So yeah, it’s amazing.

Sarah: Wow. Obviously, there are so many questions that branch off from this paper. What for you would be the biggest one that you’re most excited about tackling next in your work?

Femke: Um, well, mainly, I think that what from a theoretical perspective, what interests me the most is because yeah, there are so many small questions that I’m very passionate about as well. But the main thing is this context dependence, this probable context dependency of this mechanism. So how this relative importance of the environmental heterogeneity of the local site fidelity and then the conspecific attraction, how the importance of these things, but also the mutually exclusive behaviors, how they can change from population to population. And I think it would be a hard one to tackle, because you would need a lot of data, I think. Why I think in general, it’s the thing that is very interesting to look at, because also, with climate change, there are so many factors that change for populations that we don’t know, like the increase or decrease in population sizes, environmental factors, and the patterns that we think we always see, they might just be disrupted out of nowhere.

Sarah: That’s a really hard thing I think about working at the population level is everything is so context dependent. And when you study a single population, you get answers related to one context. But you don’t know anything about how those might apply to other populations, whether they’re responding to the same context, whether their circumstances are completely different.

Femke: Yeah absolutely. The hardest thing is that, for instance, for my model, you need really quite detailed data. And if you then want to extend this to other populations, because yeah, this question of this context dependency, it’s everywhere in ecology, right? But yeah, then you need a huge amount of data or maybe, yeah, maybe I just have to look further into like other possibilities. And maybe I will come up with an idea to combine things or get an answer to the questions in a different way, where I need less data, but still, yeah, I don’t know. I still don’t know how to have to figure it out. We’ll see.

Sarah: Yeah. Definitely lots of work still to do. But such a cool system and such a cool way to approach these problems.

Femke: Yeah. Thank you so much, also, for your interest. I’m really glad it was picked up. Like I said, it’s like, I’ve studied this since my master’s since 2016. And it’s really been like, I also I think that’s often that you are not aware of this when you see a paper and then you’re amazed by it. Oh, wow. But then, you know, it’s been a hard travel for several years, for four or five years. This is my main paper for my PhD. Of course it was this big, big process as well. And also I had a lot of feedback from the reviewers that also improved it a lot. And truly yeah, so people only see what comes out of it, of course, like the final product. There’s so many things behind it, of course, and a process for myself to work on this story.

Sarah: Yeah, absolutely. Well congratulations. It’s really an amazing piece of work and a very cool dissertation project!

Femke: Thank you so much. Thank you.

Thank you so much, Femke, for sharing your time and insight with us. Thank you to Associate editor Benjamin Bolker and the two reviewers of this paper for their work in helping bring Femke’s science to the public. Thank you to Daniel Nondorf for composing our beautiful intro and outro music. And thank you for listening! If you’re curious to learn more, and I’m sure you are, go read Femke’s paper! Send her more questions! Papers are meant to start conversations. I’m grad council rep Sarah McPeek, here to keep the conversations going.