157 - Katie Arstingstall - Bees collect pollen from more plants than you think (in English)

Andony Melathopoulos: [00:00:00] The holy grail of restoration ecology for pollinators is to have a plant list that encompasses all the interactions between pollinators and the plant community that existed before disturbance. Now it's easier said than done. You can look for bees and make these point observations of where bees are on flowers. But my next guest is going to make the case for a more comprehensive approach, that combines both genetic techniques and also detailed observations of where plant communities are. Katie Arstingstall is a wildlife biologist working for NCASI, the National Council for Air and Stream Improvement, doing pollinator work here in Oregon.

[00:00:42] But before that, or like earlier this summer, she had just completed her Masters of Science in Wildlife Science at Oregon State University, working with Dr. Sandy DeBano out in Hermiston. Katie's work revolved around this really wonderful technique, known as DNA metabarcoding in which she would take [00:01:00] pollen from bees and identify where they've traveled through the landscape. So in this episode, we're going to hear about that technique. But also she's going to tell us how it can be applied specifically in the context of Eastern Oregon in better understanding bee communities, plant communities and their interaction.

[00:01:19] That is going on all across the planet right now, the home office juggle. Now listeners of the show know and I'm sure we've heard this before that there's a tight connection between pollen, bee nutrition and plant reproduction. But we've also all observed different bees being choosy about the kind of flowers they visit. To begin, can you tell us about some of the challenges that are associated with figuring out the pollen preferences of different bee species?

[00:01:53] Katie Arstingstall: [00:01:53] Yeah, sure. So I guess when we're out watching bees visiting [00:02:00] flowers, or trying to collect them with nets we're limited in a few major ways. So typically as bee researchers, we're trying to describe interactions between bees and plants in pretty large areas and over pretty long timescales. But one person, or even like a team of people can only cover so much area and they usually have a pretty limited amount of time as well. So ideally it'd be awesome if we could follow a bee from the moment it leaves the hive or its nest to the moment that it returns after that full foraging bout. 

[00:02:39] But that's really impossible because bees can travel several miles from their nest or their hive. And so we're typically observing an area that's just like a few square meters when we're researching bees. So we end up observing just this really small snapshot of a bees foraging behavior. [00:03:00] So you could be watching a bee go from one lupine to the next and think, "oh, this bee really loves lupines!"

[00:03:09] Yeah, but then who knows? I mean the next second it flies a hundred meters to a patch of bull thistle and goes to town over there. So yeah, it's just really about being like limited, like physically and by your field of vision. It's really hard to get that full picture. 

[00:03:26] Andony Melathopoulos: [00:03:26] You know I've often thought along those lines that there's times of the year when researchers are out and then like, everybody's in class - and we don't really know what happens in the spring. Which is a slightly different issue, but there may be the resolution of a plant, bee connections may vary for a lot of different factors. But if I understood you correctly, what you were saying was that if you're going out, you catch some bees and you see them on a specific flower, it may move around and you're really just getting these [00:04:00] very tiny snapshots. So we're kind of handicapped to a certain extent of being able to know all the flowers that they would visit in a forging trip or over their entire lives. 

[00:04:11] Katie Arstingstall: [00:04:11] Right, exactly. 

[00:04:12] Andony Melathopoulos: [00:04:12] Okay. On a previous episode, we had Pierre Lau from Texas A&M who talked about two methods that maybe could help with this - how to identify pollen on a bee's body, either using microscopic analysis or pollen DNA. And I guess if we could see the pollen on their bodies it would maybe give us a broader picture of where they've been. Can you remind us of some of the benefits and limitations of these different methods of analyzing pollen on the bee's body?

[00:04:42] Katie Arstingstall: [00:04:42] Yeah, totally. So microscopy is this method that a lot of pollinator scientists use where you look at the pollen grains under the microscope and you can separate them out based on their physical differences to see what plants [00:05:00] species they come from. And so this method is really cool because it helps resolve a lot of the issues with direct observation. So we're no longer, really limited in time and space. All the pollen that, that bee collected is right there. So as long as we're able to identify those, then we really do get that like "full picture" it's no longer really this snapshot anymore. 

[00:05:27] But there are some limitations to this method as well, microscopy is typically as I understand it, you only use a subsample of the pollen load. So it is homogenized beforehand so it should be a pretty good blend of what that bee was taking from. But you still could maybe miss some rare species that were in there. 

[00:05:56] Andony Melathopoulos: [00:05:56] So just to cover this, so you would get the bee [00:06:00] you would sort of knock the pollen off its body and you'd have to mix it somehow because the chamber for counting the pollen may be only microliters in size. So you're only getting this small fraction and you're making this assumption that you mix it up. But if there's just a little bit of pollen in there, the odds are as you might miss it. That's what you mean by rare species might get missed. 

[00:06:21] Katie Arstingstall: [00:06:21] Right, exactly. 

[00:06:22] Andony Melathopoulos: [00:06:22] Okay. Got ya. 

[00:06:24] Katie Arstingstall: [00:06:24] Yup. And then another limitation is that the taxonomic resolution for this method is typically pretty low. So what I mean by this is that you're probably only going to get like a family or genus level identification as opposed to specific species. And the species level identification is really important when you're looking at plant, bee associations. So yeah, from what I've [00:07:00] understood from people who use this method, you can only really ID pollen grains to that higher taxonomic level. I've heard it's really, really difficult. 

[00:07:10] Andony Melathopoulos: [00:07:10] Yeah. It seems like there's a couple of very good specialists who are overwhelmed. But most people can't do it - it's a very challenging skill. Okay what about this other method that you don't have to use a microscope and actually physically look at the structure of the pollen? How does that work? 

[00:07:31] Katie Arstingstall: [00:07:31] Yeah, so in the past, I'd say like five years or so it's become really popular. A lot of researchers have started using this technique, it's called DNA metabarcoding. So this allows you to take the pollen from the bee and sequence the DNA. We can use the DNA sequences to identify the species in a bee's pollen [00:08:00] load. So instead of determining the species based on their physical appearance, you're using the actual DNA.

[00:08:10] And so I know Pierre explained how all plant species produce different pollen grains. And you separate them out based on their morphological differences. Well, similarly, we can tell species apart based on their DNA. And so I'm sure that you've heard, like as humans, we share 99% of our DNA with chimpanzees and that 1% is what makes us different species. And that's super similar with plants in that they share the majority of their DNA as well, the different species. But there's a small amount of DNA that makes them different species. And so we can use these specific gene regions that are [00:09:00] different between different species to separate out the plant species in a pollen grain.

[00:09:08] And so the way this works is you take the pollen load, you can take the pollen load off of an individual bee. And this might contain several different species of plants that it's collected from during that forging bout. And we extract the DNA from the pollen then we use a PCR which stands for polymerase chain reaction, to amplify this specific gene region that is going to help us tell the different species of plants apart. And so then once we amplify these gene regions we are able to download from the internet known [00:10:00] DNA sequences, all kinds of plants. There's a ton available online, and then we can essentially just match the DNA sequences from our pollen samples, to those of these using known plant DNA sequence. Then we can figure out what plant species are in the pollen loads from these bees. 

[00:10:23] Andony Melathopoulos: [00:10:23] This sounds like it potentially has more power than looking at the pollen grains under a microscope. Are there any downsides? 

[00:10:38] Katie Arstingstall: [00:10:38] Yeah, there are definitely some limitations to this method. One is, there's a really big learning curve when it comes to understanding the different genetics lingo and learning all the techniques involved with the lab work. As well as there's a lot of coding involved [00:11:00] which was all new to me when I started this project.

[00:11:03] Andony Melathopoulos: [00:11:03] Because there's so much data that you produce, you have to somehow congeal it. 

[00:11:14] Katie Arstingstall: [00:11:14] Yeah, a hundred percent. Yeah, so it's a beast to get through all of the information, but another limitation of this method is it's not quantitative. So basically we can tell what plant species are in a pollen sample, but we can't tell how many of each are in there. 

[00:11:40] Andony Melathopoulos: [00:11:40] Oh, and I guess coming back to your question of like, you know, “the person who wants to know which plants are the most important for which bees." And you set it up real great at the beginning. As you know, we only see them on one plant, but if at the end of the day you get all the pollen off them and it tells you the list. Maybe they only kind of briefly [00:12:00] visited it. And really that one plant is not that important, but it will look the same as all the plants, because it'll come up as a kind of inventory of where they've been or something.

[00:12:10] Katie Arstingstall: [00:12:10] Right. There are some ways that we can like, kind of try to get around this. So like for my project, we set up these transects and our sampling areas and we basically would walk along these transects and count the number of blooming stems from each plant species that was growing along those transects.

[00:12:34] Andony Melathopoulos: [00:12:34] Yeah. 

[00:12:35] Katie Arstingstall: [00:12:35] So this gives us an idea of the presence and abundance of what's blooming in the area. And so then you can use this data to essentially say, "okay, well, let's say bull thistle we only saw it like 5% of the time on the transects, but we found it an 80% of the pollen loads." Then you could, you could at least assume that plant [00:13:00] species is somewhat preferred by these bees. 

[00:13:03] Andony Melathopoulos: [00:13:03] Oh, because there's not a lot of it out there. And it shows up on every bee is kind of like probably trying to get to that one bull thistle. And so that way, you know it's significant. That makes a lot of sense. So although the tool may not give you the quantity, you can sort of infer it by walking the landscape and matching the two data sets together.

[00:13:24] Katie Arstingstall: [00:13:24] Yep. 

[00:13:25] Andony Melathopoulos: [00:13:25] Okay, fantastic. I have one more question about this before we take a break, then we'll dive into your research because I think we're already kind of like dipping our toes in your research. I'm not genetics person at all, but you're talking about regions of the DNA that you look for - does every plant have different regions or there's like one or two regions that you focus on? I'm thinking of it like a roadway there's just like mile 34 to mile 78 is where we're [00:14:00] looking. Is that the case or is it more way more complicated than that? 

[00:14:04] Katie Arstingstall: [00:14:04] That's a really good way to look at it. It's essentially, so they have names like the two that I used in my project, one is called RBCL and this is one found in the chloroplast DNA.

[00:14:23] Andony Melathopoulos: [00:14:23] Okay. 

[00:14:24] Katie Arstingstall: [00:14:24] So really good for plants. We actually use two gene regions for my project because that actually gives you a little bit higher taxonomic resolution. You're able to identify more using two than one because sometimes plants might have the same RBCL region, but different ITS2 regions, which is the other gene region that we used which is actually found in the nuclear DNA.

[00:14:58] So yeah. [00:15:00] Lots of people that are much smarter than me, have done tons of research beforehand and figured out certain regions that are good for certain organisms. So there's some that are really good for plants, there's some that are really good for like fish. There's some that are really good for other organisms.

[00:15:24] Andony Melathopoulos: [00:15:24] Okay. Fantastic. Well, thanks so much. Let's take a quick break. I want to come back after the break and talk about your specific research. Because it sounds like you've done this really wonderful kind of blend of classical ecological work with DNA work to be able to get at the central question that we raised at the beginning of like, where do these bees actually go?

[00:15:47] Katie Arstingstall: [00:15:47] Yeah, definitely. 

[00:15:48] Andony Melathopoulos: [00:15:48] Okay. We're back, alright. So we were talking in [00:16:00] the part before the break, how you've really tried to refine and kind of made this bridge between these genetic methods and actual observation on the ground. You know, sort of refine these existing methods by first developing a library of plants that are actually growing in the area, rather than this huge library that you were describing on the internet, where there can be plants from Madagascar and plants from England. You thought, well, one way I can get this whole system to work better is by really focusing on what's the flora of that area. Can you tell us a little bit about how you went about doing that? 

[00:16:35] Katie Arstingstall: [00:16:35] Yeah, definitely. So as I was mentioning when I was explaining the process of DNA metabarcoding once you have your DNA sequenced from the pollen, you can match that DNA with what we call a reference library that contains known DNA [00:17:00] sequences that people from all over have added to databases on the internet. We wanted to build a library that essentially just had plants in our area because certain families and genera can share a very significant proportion of their DNA at these gene regions that we are using to identify species from. 

[00:17:41] For example two kinds of knapweed, spotted knapweed, and diffuse knapweed. They differ in their ITS2 gene region, which is one of the ones that I used by only one base pair out of 500. And so [00:18:00] sometimes our pipeline can mix them up and can mis-identify one as the other. And so when we were building this library my advisor, Sandy Debano has been working in my study areas for quite a few years. And she's been working with several botanists and other researchers at the Nature Conservancy and the Forest Service, and they've developed a pretty comprehensive list of the plants that grow in these areas. 

[00:18:35] So we were like, "oh, awesome, like let's just take these lists and kind of develop our library based on that." So when we were creating this first, we checked curated databases which are essentially these like really top notch databases, where sequences have been [00:19:00] verified by botanists. So you can be like super sure that everything's very accurate. And so you can find the links to these kinds of databases and different journal articles. That's where we found the ones that we use. 

[00:19:15] And then everything that could not find from those curated databases we downloaded from an open source database called GenBank which essentially anyone can sequence DNA and add it to GenBank. It has to get accepted and go through an application process.

[00:19:40] Andony Melathopoulos: [00:19:40] So you had two tiers. You wanted to go with the really verified sequence data first, and then if there was something missing you would use the second one. Can I just ask you a quick question on this? I imagine the amount of plants that bees visit in the study area you're working on, were hundreds. So [00:20:00] you had to go through like line by line and then it's like "this plant, where's the best sequence?" That sounds like a lot of work. 

[00:20:10] Katie Arstingstall: [00:20:10] Yes. So that's where the coding comes in. And so I had this really awesome colleague, his name is Xiaoping Li. He is a postdoc with Ken Frost in the department of Botany and Plant Pathology at OSU. So Xiaoping actually wrote up a little script that you basically could just run it and it would essentially check what was in the database and give you back like ones and zeros if it was there or not there. So yes, by hand that would have taken forever. 

[00:20:50] Andony Melathopoulos: [00:20:50] So you started with this great data set of actual plants on the ground, and then you had to assemble a corresponding database of sequences using this [00:21:00] fancy super program. And then how did you proceed from there? 

[00:21:07] Katie Arstingstall: [00:21:07] So I guess on average we were able to download DNA sequences for about 70% of the plants that were in our plant lists which is pretty good amount. And so then essentially we run another pipeline to match those sequences. And then, like you were saying before, it essentially spits out a list of what plant species were found in each pollen load. One thing we wanted to do was to compare this data to actual [00:22:00] behavioral observations like we were talking about in the beginning. So as we were catching these bees we were noting the flowers that they were visiting while we were sampling them. So we can essentially directly compare observation data to our metabarcoding data to see if there were any differences. 

[00:22:31] Andony Melathopoulos: [00:22:31] That sounds exciting. Okay. And what was the answer? This is how you set things up, like this is the way we've been doing it for a long time. When we say, "oh, this bees's a specialist bee" it's maybe based on a couple of observations of the bee on a flower. So when you lined up the proof in the pudding, the actual genetics with the observational data, what did you see? Where they're visiting a lot more plants than we might've thought they were. 

[00:23:01] [00:23:00] Katie Arstingstall: [00:23:01] Yes, they were. Basically we found that their diets are a lot more complex than what you would assume based on just direct observations. For example, at Starkey Experimental Forest and Range, was one of my field sites in Eastern Oregon. We documented 26 plants species that bees were observed visiting but when we looked at the metabarcoding data, we actually found 44 plant species in those pollen loads. So almost twice as many species. And not only were we detecting more species, we were detecting more interactions per species. So each bee's pollen load contained anywhere from one to nine different plant species. And so, like you [00:24:00] said, a lot of these bees that appeared to be specialists based on just observation now looked like they had more of a generalist diet. So it's super interesting. 

[00:24:12] Andony Melathopoulos: [00:24:12] Well, and I think just as a first point as well that when we go out and we do these surveys of bees and the plants they visit, we may be underestimating by a lot, the number of species that are actually being visited. You mentioned earlier, this kind of, you know, going through and seeing what plants were actually out there. Was anything surprising? Where there plants that were sort of performing much differently than you would have thought given just doing observational work?

[00:24:50] Katie Arstingstall: [00:24:50] Yeah, definitely. One in particular that I think of is yarrow. At [00:25:00] least in my experience, I don't see bees on yarrow very often. But it was probably like the third, most common species in the pollen loads at two of our locations, which was really cool. Another one that comes to mind is rabbitbrush. So I had some field sites at a large industrial farm. We sampled in the field margins and there was a lot of rabbitbrush just not in our plots. So we randomly selected I think eight field margins to sample from. 

[00:25:49] And so there was a lot of rabbitbrush in the surrounding area, just not where we were sampling it. And so I had no observations of these visiting that species. But [00:26:00] it was the second, most common species in the pollen samples. So, I mean, that was like a perfect example, we never would have known that they were even using that if we had not done metabarcoding. 

[00:26:15] Andony Melathopoulos: [00:26:15] I love the thing that you said at the beginning, like most researchers work on a plot and like bees don't work on a plot. Because you may be doing some kind of restoration on a plot and you're really focused on it, but your bees may actually be hanging out in the plot, but they're doing a lot outside the plot and you would never know. 

[00:26:37] Katie Arstingstall: [00:26:37] Yeah. It was a lot of fun looking at the data. I also remember elephant's head being in the pollen, which is that really cool, like spike hot pink flower, that looks like an elephant head. And I remember seeing that it's Starkey like one time. But [00:27:00] that was showing up in there. Just lots of cool stuff that I never would have even known that they were using.

[00:27:05] Andony Melathopoulos: [00:27:05] Why don't you kind of take us through it a little bit. It's a part of the state that I think has some remarkable bee biodiversity. Many of our listeners may have not been out there. Tell us about some of the other cool plants that popped up through your analysis and some of the bees that you ran into.

[00:27:25] Katie Arstingstall: [00:27:25] Yeah. So I had three different study locations all in Eastern Oregon, they were all relatively similar to each other, but very different ecosystems. So I worked at the Zumwalt Prairie Preserve, which is a remnant bunch grass prairie. It has like a very rich forbe community, has a lot of flowers that are typically found at higher elevations. Like, Clarkia's, [00:28:00] paintbrushes. There is some larkspur lots of lupine and the bee communities out there were pretty diverse. We sampled about 35 species from Zumwalt and I would say it's pretty dominated by bumblebees or at least that was the most of what we caught. Specifically we had a lot of Bombus californicus and Bombus centralis. And then we had a mix of all the other genera as well. Lots of Andrena, Halictus, Lassioglossum. But the bees at Zumwalt definitely loved the lupine and that's been my experience like everywhere I go typically, is if there's lupine, then the bees are going to find it. The bumblebees love it.

[00:29:00] [00:28:59] And then my second field site was at Starkey Experimental Forest, which had several forested riparian sites there. So my sites ran along Meadow Creek which is a tributary of the Upper Grand Ronde River. And so there's also a higher diversity of flowering plants and bees at Starkey as well. As far as plants the Zumwalt and Starkey overlap quite a bit, but at Starkey I would see more Missouri goldenrod, St. John's-wort, penstemons also really common. And we actually sampled individuals from 40 different bee species there. Again, a lot of bumblebees at Starkey we collected a ton of Bombus bifarius we also caught quite a few Halictus, [00:30:00] Lasioglossum species as well. 

[00:30:05] Then my last site, Three Mile Canyon Farms, which is I would say pretty disturbed landscape. 

[00:30:13] Andony Melathopoulos: [00:30:13] This is in Eastern Oregon, but whereabouts is it?

[00:30:17] Katie Arstingstall: [00:30:17] It's right next to the Boardman Preserve it's in Morrow County. Yeah, really close to Hermiston where OSU has their extension center. It's a pivot irrigated farm, they keep their field margins uncultivated. 

[00:30:39] Andony Melathopoulos: [00:30:39] I've been there. It's that place that does all the international hay exports. 

[00:30:46] Katie Arstingstall: [00:30:46] Sure, yeah. I didn't know that.

[00:30:50] Andony Melathopoulos: [00:30:50] That was also the site that you described earlier when you were talking about the plots and the rabbitbrush. 

[00:30:56] Katie Arstingstall: [00:30:56] Yes. 

[00:30:56] Andony Melathopoulos: [00:30:56] Okay. Alright. How did the bees look out there? 

[00:31:01] [00:31:00] Katie Arstingstall: [00:31:01] Very different. The bee communities and the plant communities were very different at Three Mile. As far as plants, lots of no native vegetation. There were several knapweeds and Russian thistle and Scotch Thistle. And as far as bees, less diverse, as far as what we sampled. We sampled 24 bee species, but I was still really actually, I was very surprised at how diverse it was there because when I first started sampling there, I was like, "oh, it's going to be all honeybees, I'm not gonna find any native bees." But there were a lot of super pretty Agapostemon, green metallic bees lots of digger bees, they love that soil, several Mellisodes species. I'd say a lot less bumblebees [00:32:00] at Three Mile.

[00:32:01] Andony Melathopoulos: [00:32:01] It's something that I've noticed working in the Dalles this summer. Just after you leave Hood River and you head to Hermiston it's just not a great bumblebee area. But then you go up in elevation to Starkey and Zumwalt, then it suddenly is. But it just strikes me that whole stretch is just not great, I don't know if it's too dry, it's too early, there's not enough late bloom or something, but I've noticed the same thing. 

[00:32:29] Katie Arstingstall: [00:32:29] Yeah, it's kind of like a desert there for a while. 

[00:32:36] Andony Melathopoulos: [00:32:36] Well, those are three really different sites. So that must've been a really good confirmation of this approach because if you were able to have better correspondence between your genetic data and the actual plant community that would be a good [00:33:00] test of it. You would be looking at a real broad range of landscapes. 

[00:33:05] Katie Arstingstall: [00:33:05] Yeah, definitely. And we haven't started writing this paper yet, but we are going to look at several factors comparing the bee communities and the plant communities between those very different landscape types. And what we found is that it seemed like the bees at Three Mile were exhibiting a lot more of a generalist diet. So it's almost like they're just making do with what they have out there, using a little bit of everything. Whereas at Starkey and Zumwalt where we have this high plant diversity there's a lot more room for specialization is what it looks like.

[00:33:48] Andony Melathopoulos: [00:33:48] Oh, that'd be a really great finding to document. I think that's really interesting. Well, let's take a quick break. We'll come back, we have our questions that we ask our guests. [00:34:00] But let's take a quick break and we'll come back. 

[00:34:03] Katie Arstingstall: [00:34:03] Sounds good. 

[00:34:05] Andony Melathopoulos: [00:34:05] Okay, we're back. So first question is do you have a book recommendation for our listeners? 

[00:34:12] Katie Arstingstall: [00:34:12] Yeah, so I can almost guarantee somebody else has recommended this, but it's so good. "The Bees in Your Backyard" is my favorite. Just the pictures are amazing, but it's also full of so much awesome information about native solitary bees. It goes through all the genera specifically, so I really love that book. 

[00:34:37] Andony Melathopoulos: [00:34:37] I remember seeing it for the first time at the University of Calgary, Lincoln Best showed up with a suitcase of them. And he said, "you guys all need this!" But the other thing I do know is that we were talking to Olivia Messinger-Carill, one of the co-authors and I think they've sent to the publisher, "The Bee's of Eastern US", and they're working on a "Bees of Western US" so [00:35:00] same really great images and lots more detail. I'm so excited to see those. 

[00:35:05] Katie Arstingstall: [00:35:05] That's awesome. 

[00:35:06] Andony Melathopoulos: [00:35:06] Great recommendations, also my favorite too. Okay. I imagine you do a lot of different things, you've had to master a lot of esoteric techniques and technology. So I'm really curious at what your answer is going to be to this question. What is your go to tool for the kind of work you do? 

[00:35:31] Katie Arstingstall: [00:35:31] Yeah. So I had to think about this one, because my tools are pretty spread out. But, one thing that is so important in identifying plant pollinator relationships is plant ID. If you don't know what you're looking at out in the field and you're not going to get as great of data as you want. So I picked, "The Wildflowers of the Pacific Northwest" it's a [00:36:00] field guide by Mark Turner and Phyllis Gustafson. It's organized by color and then by shape. So it'll have like yellow flowers and then it'll say like three pedals, four pedals, five pedals. So it's super like user-friendly, even if you don’t have any experience in plant ID it's really great, it's got awesome pictures!

[00:36:24] Andony Melathopoulos: [00:36:24] Fantastic. We'll put a link to that in the show notes. It's really really vulgar, but there's a really great YouTube channel called, "Crime Pays Botany Doesn't". The guy who does it, he's some kind of train conductor from the Bay Area who travels all across the West and he does botany. Because I'm the same way, I like the "Oregon Flora Guides", you know, the [00:37:00] app with the flower shape and then the pedal colors. But he really did emphasize to me getting a basic book on, basic botany and taxonomy, how the plants are organized. 

[00:37:12] I've never really done that, I know these plants are related, but I imagine you've had to do that a lot, that you now have a real intricate knowledge of how these plants are related. Because one could be pink and one could be white, but they could be basically the same plant. I'm trying to try and learn systematics of plants this winter! Yeah. Okay well, that's a great recommendation. I'm always looking for a great field guide. It'll be in the show notes. 

[00:37:43] And the last question we have, it's always an unfair question for somebody who has probably has a lot of different pollinators that they love, but is there one in particular when you see it fly by, you're like, "ha I love that thing"?

[00:37:55] Katie Arstingstall: [00:37:55] Yeah, I do have one the scientific name is Anthophora [00:38:00] curta the common name is short sun-digger. It's a digger bee, it's just so cute and it looks like a little teddy bear. It's got these like teal eyes and like light blue hair. I just think it's adorable. Really furry legs. I love it! 

[00:38:20] Andony Melathopoulos: [00:38:20] What was the species?

[00:38:22] Katie Arstingstall: [00:38:22] Anthophora curta.

[00:38:25] Andony Melathopoulos: [00:38:25] It's not an Oregon Anthophora, is it? 

[00:38:28] Katie Arstingstall: [00:38:28] I'm not sure. It was all over at Three Mile Canyon farms. 

[00:38:37] Andony Melathopoulos: [00:38:37] Oh, it is. Okay. I have never seen a bluish one. That's really cool. 

[00:38:42] Katie Arstingstall: [00:38:42] Yeah. It's light blue and awesome!

[00:38:49] Andony Melathopoulos: [00:38:49] Fantastic suggestion. It would be really great to have a blue Anthophora for I'm going to now hunt for the blue Anthophora. 

[00:38:58] Katie Arstingstall: [00:38:58] And they're super hard to catch, they're [00:39:00] extremely fast. 

[00:39:07] Andony Melathopoulos: [00:39:07] On larkspur I remember seeing them and it was just like, you had to sit for a long time and plan your collection. Because you were never going to catch it. It's sort of like, "it's going to come back in three minutes, I'm going to wait." This is fantastic. Thank you so much. I think this is really kind of filled out our understanding of these complicated connections between bees and pollen. Good luck with writing up those papers. We're really excited! Maybe we'll have you back after it's published so we can discuss them.

[00:39:40] Katie Arstingstall: [00:39:40] Yeah, that'd be great. Thanks for having me.