Convo with Elsa Youngsteadt: How urban heat affects bee populations

elsa youngsteadt GRAPHIC B


Here’s a recent conversation I had with Elsa Youngsteadt, a research associate in the department of entomology and plant pathology at NC State. She and her colleagues, April Hamblin and Steven Frank, recently published a paper showing how urban heat affects bee populations. The transcript below has been edited for length and readability.

Matt: Before we start talking about the paper, can you give me a little background on how you got into bees?

Elsa: Sure. There was a period after I finished my PhD and before I started my current position where I was quitting science and being a science writer and editor. I worked for a few years with American Scientist magazine and worked with PRI on a science podcast and did some freelancing. And that’s when bees got on my radar because, of course, they’d been in the news a lot over the past decade with declines in bee health or certain populations of bees. So like a lot of people I got concerned about that, and I thought, “Well, I guess there’s something I could do as a researcher to take an active role in understanding what’s going on.” So when I came back as a researcher in 2012, that was one of my goals.

It’s not that I’m exclusively a bee researcher — I study the ecology of insects in urban environments more generally. I’ve worked on ants and scale insects. And bees are another member of the insect community that lives in cities.

M: What in that sequence of events got you looking at the impact of heat on bees?

E: Yeah, we’d been looking at the effect of heat on other insects besides bees in cities. It started with work on scale insects. We’d been finding that scale insects actually benefit from urban warming, and they become much more abundant and pesky around street trees in urban hot spots. But that was really focused on just two species, and so the question was if there are two species of insects in cities having such a dramatic response to warming, what’s going on with the rest of the insects?

Bees are a great group to look at that question with because they are so diverse in cities. We’ve logged more than 100 species just in our urban area around Raleigh. So you can really get into questions of, “Do different species respond differently to warming? Do some benefit while others decline?”

M: Right. Can we backup for a second? You said you started by looking at two species of “scale” insects? What are those? I’m sorry I don’t know.

E: (laughs) No problem. They’re kind of obscure. They don’t move much so you don’t really notice them. They’re all herbivores. They have this straw-like sucking mouth part that they tap into a plant and drink the fluids. They just sit there and do that for most of their life. Then there are very brief periods in their life cycle when they actually move around. So that’s why they’re called scales: they look like little scales on a tree or on a plant. When they become very abundant they can do a lot of damage. So red maples are one of our most common street trees in Raleigh and they host a very common scale insect called the gloomy scale. When this scale becomes more abundant, the branches start to die back and the tree starts to look really scraggly and unhealthy. And this happens more in the hotter parts of the city than in the cooler parts.

But originally there wasn’t much explanation for why some trees really suffer and other trees do just fine a block over. And that’s why we started looking at the urban warming question because we know that heat can vary from one end of the block to another. So that fine-scale temperature variation drives these differences in pest populations on street trees, and there was a really strong relationship between the temperature at the site where the tree is growing and the scale insect infestation.

M: So what did you discover about the relationship between urban heat and bees in this recent paper? Maybe give us the elevator-pitch version.

E: It’s hard to say what we found without saying what we did, so this may not be a good elevator pitch…

M: You can pretend it’s a long elevator ride.

E: (laughs) Given this background of what we’d seen urban warming doing with scale insects, we wanted to know how it was affecting bee communities in Raleigh. And this was a project that was led by April Hamblin during her masters studies. She’s the one that did all the sampling and the fieldwork. She selected 18 sites around Raleigh, some warmer sites and some cooler sites, all in residential areas or parks. Over two years she sampled the bee communities in those sites 11 times, collecting bees using a net and a couple different kinds of traps. Then she asked how the abundance of bees and the diversity of bees varied because of the temperature at the sites where she was sampling.

HamblinPrairieRidge 960x
Photo courtesy of April Hamblin

She also measured other things about the sites that we know should matter to bees: things like the abundance and diversity of flowers, the amount of impervious surface like pavement and buildings surrounding the site. So putting all of that together, we found that temperature — as well as impervious surfaces — were important predictors of bee abundance. And looking at the bee community as a whole, their abundance declines by about 40 percent per degree Celsius in urban warming. The hottest spots had the fewest bees.

But not all species declined at the same rate. April encountered more than 100 species in her samples. Some had fairly steady populations regardless of the temperature, whereas others declined very steeply. And that’s what led into a previous paper looking at how we can understand which bees are going to be the most sensitive, therefore predicting which ones will be tolerant versus which will decline.

That was a long elevator ride…

M: (laughs) No worries. So I found it curious that bee abundance is impacted by temperature but species richness is not. Only because both in the paper and just now you said some species were impacted more than others by temperature.

E: Yeah, so species richness is kind of a funny measure. It’s not just the raw count of the number of species you find. It’s adjusted for the number of bees considered in that sample. So, for example, if you find ten bees at a site and ten species represented, versus if you find 10 bees and only one species represented, then that’s an obvious difference in species richness. But if you have a sample of 100 bees from one site and only ten bees from the other site, well then how do you compare that? Because, of course, there are going to be more species in the 100; you just have a better chance of detecting things. We actually want to take that effect out of the results in thinking about species richness. So you re-sample your 100 bees back down to ten and then make a direct comparison. And these are just example numbers, we’re not really talking about samples of exactly ten and exactly 100.

M: So is the non-impact on species richness just an artifact of the way you compare apples to apples?

E: I mean, if you did go to the hot sites and just keep sampling until you had as many bees collected as you had from a cool site, the pattern in the data so far suggests you would accumulate as many species. It would just take you much longer because there are fewer overall bees. They might not be the same species but…

M: Interesting.

E: The more important difference is the decline in the total number of bees versus the change in the species composition. Even though we didn’t find differences in the diversity given that correction for sample size, we did find differences in the species composition of the communities. So if all bees are a pie, some bees get a bigger slice of the pie at hot sites than they do at cold sites and vice versa.

M: Right. So what you’re saying is that while the count of specific species might change between sites, the overall number of species in the community composition at each of those sites was comparable and didn’t change based on temperature.

E: And the thing that’s driving the change in composition is the fact that some species decline faster than others. You might have a very heat tolerant species with similar abundance at a hot site and a cold site, but another species that’s very abundant at a cold site and becomes rare at a hot site.

M: And what are the common factors related to a negative relationship with an increase in heat?

E: So, very few bees appeared to actually become more abundant with warming. It was more like, “Who declined the least?” It would make sense that a bee’s actual physiology and thermal tolerance would predict whether or not it would remain abundant versus declining at hot sites. And our previous paper looked at this issue.

April did a physiological assay with 15 species of bees where she measured their thermal tolerance, known as the CT max. That’s basically the temperature at which a bee goes into a kind of stroke situation, where it falls over and doesn’t really function any more. She brought bees into the lab, warmed them up and determined the sort of heat tolerance marker for those 15 different species. The idea being that the ones with the highest heat tolerance in the lab should be the ones with the stable populations across the city. Whereas the ones with the low heat tolerance in the lab should decline at hotter sites. And that is, in fact, what we found.

M: Looking at the paper, body size was not a common factor. But solitary bees, and ground-nesting and stem-nesting bees were the most impacted. Does that sound about right?

E: Yeah, so solitary bees were more sensitive overall than social bees. They tended to have a lower thermal tolerance and social bees had a higher thermal tolerance. But then, kind of ironically, social bees include both ground nesters and cavity nesters. And by “cavity nesters” we’re really referring to bumble bees, because that’s the only kind of bee in that category. And bumble bees were more sensitive than ground- or stem-nesting bees. So depending on how you slice up the bees you could get a different answer. Because although bumble bees are social and also cavity-nesting, bumble bees as a group were quite sensitive and had a lower thermal tolerance compared to other groups of bees. But when you include bumble bees in the social category, overall solitary bees tend to be more sensitive.

Photo courtesy of April Hamblin

M: So here’s a question I bumped up against while reading through all of your work: The heat island in a city is very much related to the amount of pavement you have in a city. The impermeable surfaces are the reason why these heat islands occur in the first place. We know that solitary ground-nesting bees are impacted by heat, but how did you determine that it wasn’t just the amount of pavement decreasing the number of these bees?

E: That’s always a tricky thing in urban studies, when we want to understand this effect of heat versus pavement. Because they are correlated. And really you have to look at each one separately and say, “Which one is the better predictor of the thing that we’re seeing?” They’re not so correlated that they’re completely identical. So in our most recent work, each was similarly good at predicting bee abundance. Temperature was a little bit better of a predictor.

Also, it seems that if you’re finding this correlation between physiological effects of temperature on the bee in the lab then that supports the idea that it’s a mechanism by which temperature is affecting them in the field.

M: Because you already demonstrated in the lab how heat impacts different species differently.

E: Yes. And then when that same pattern plays out in the field, you can feel pretty good about saying that temperature is what’s driving the pattern. Even though impervious surface kind of goes along with it.

But results could differ depending on your city. So Raleigh already has a lot of green space. It’s not a very tightly packed city. I think our highest measure of impervious surface was maybe 40% or 50% surrounding a site. Which in some cities is the other end of the spectrum: that’s your least impervious site and it goes up to 90% or something. So at some point you might expect to see a transition where there really is no ground left for these bees and they cannot nest. The effects of temperature might get lost because some other resource becomes limiting first.

In Raleigh, we felt pretty good that these effects are related to temperature because we can do the temperature assay in the lab, we can see it playing out on a larger scale in the field, and it seems like some of these other factors are not becoming limiting first.

M: And it strikes me that even if you can’t tease out temperature versus pavement precisely, temperature is still a factor that hasn’t been thought about before.

E: Exactly.

M: So what’s the next step in exploring this new understanding about the impact of the urban heat island on an urban bee community?

E: There’s something that I’m working on now that kind of branches off from this. We’ve shown the direct effect of warming on bees through their physiology. But of course they’re also eating pollen and nectar, they have to have food from flowers. So how does the warming affect the plants themselves? Are warming and urban air quality affecting the quality of the flowers, the pollen and the nectar available to the bees?

M: Similar to the study that came out in 2016 showing that rising levels of CO2 are reducing the protein in flowers that bees forage on?

E: Yeah, that’s exactly it: Cities tend not only to be warmer but also have higher CO2 concentrations than the surrounding areas because of all the traffic. So we would expect to see effects similar to that study in cities.

M: Last question for you, Elsa. For the average person, what’s the take away from your recent paper? What sort of action can it inspire? What should it get people thinking about?

E: One interesting thing we found that we haven’t talked about yet — and which is empowering to the average person doing something — is the local-scale differences. Differences in temperature that we measured on a single property or within a city block made a difference to the abundance of bees that we found. Anything that reduces the amount of impervious surface and, hence, the temperature (even on the scale of a single property or city block) could benefit bees. And even though we found that the benefits of flowers were greater only for large bees, we still did find an overall increase in bee abundance at places that had more flowers. Which is not really news but it is something that anybody can do, and for some species it probably helps counteract the effects of warming.

M: Thanks for taking the time to talk, Elsa.

E: No problem.