[Automated] This conference will now be recorded.

[Deana Crouser] Okay. Good morning, everyone. And welcome to another EcoFOCI Seminar Series. I'm Deana Crouser, co-lead of the seminar series with Heather Tabisola. This seminar is part of NOAA's EcoFOCI bi-annual seminar series, focused on the ecosystems of the North Pacific Ocean, the Bering Sea and U.S. Arctic to improve understanding of ecosystem dynamics and applications of that understanding to the management of living marine resources. Since 1986, the seminar has provided an opportunity for research scientists and practitioners to meet, present and provoke conversation on subjects pertaining to fisheries oceanography, or regional issues in Alaska's marine ecosystem. Visit the EcoFOCI webpage for more information at www.ecofoci.noaa.gov. We sincerely thank you for joining us today as we continue our all virtual series. Look for our speaker line-up via the OneNOAA Seminar Series and on the NOAA/PMEL Calendar of Events. If you've missed a seminar, you can catch up on PMEL's YouTube page. It takes a few weeks to get these posted, but all seminars will be posted there. I just want you to take a moment and check that your microphones are muted and you're not using video. During the talk, please feel free to type your questions into the chat. We will be monitoring the questions and we'll address them at the end of the talk. Now, today, I am pleased to introduce Laurel Nave-Powers. Laurel obtained her Master's of Science from Southeastern Louisiana University, where she was investigating the niche conservatism and niche packing of Cyprinids in Southeast Asia and Africa using geometric morphometrics. Currently, Laurel is a PhD student at UDub School of Aquatic and Fishery Sciences, working in collaboration with NOAA and Alaska Fisheries Science Center on the transfer of ichthyoplankton from the groundfish surveys, the UDub Fish Collection. And with that, let's begin.

[Laurel Nave-Powers] Thank you so much, Deana, for that kind introduction. So like she said, my name is Laurel, and I'm currently a PhD student in the Tornabene Lab over at UDub. And I do work on the ichthyoplankton transfer from the AFSC Archive over to UDub. So if you're around in the ichthyoplankton lab, you might have seen me surrounded by millions of tiny, tiny little vials. If you do see me, please come say hi, I would love to meet you in person. So this presentation has two different parts. We're gonna start with my Master's work. And at the end, we're gonna take a little peek into my dissertation, what I'm currently working on, but starting with my Masters. So I assessed niche conservatism and niche packing patterns in African and Southeast Asian communities of Cyprinoidei. Before we get into the real meat of my project, I do just wanna start out with some background information. Scientists have long been intrigued with the relationship between morphology and ecology. Hutchinson describes a niche as mapping population dynamics into an environmental space with both abiotic and biotic factors, influencing birth and death rates. That sounds very complicated. Essentially, it describes the role of an organism in its environment. For many organisms, morphology is a great indicator of resource use because with similar environments, you get similar selection pressures, which then, of course, translates to very similar phenotypes. For example, greater Antillean anoles have evolved repeated sets of ecomorphologies across different islands that allow them to partition their resources. Similarly, cichlids of the East African Rift Lakes, have independently evolved similar feeding morphologies to each other. So these examples are showing us how competition can drive morphological diversification through these different uses of resources and partitioning of those resources, which then minimizes competition, and then enables the overlap and coexistence of species. The ecomorphological approach states that morphology is an indicator of niche and ecological role. And we can estimate a species niche by examining different aspects of its morphology. And luckily for me, this has already been pretty well studied in fish. For example, and this is pretty basic, you probably all already have a basic understanding of this, but for example, fishes that are dorsally flattened with an upward facing or superior mouth tend to live in top water habitats and go after prey at or near the surface of the water. Fish with a more fusiform or a deeper body, tend to be in more pelagic environments and they have terminal or more forward facing mouths to go after prey typically right in front of them. Eventually, flattened species with inferior or downward facing mouths are usually benthicly oriented and go after prey at or near the bottom, the substrate. So these examples are showing how we can use different aspects of morphology and body shape to indicate niche. So while the relationship between morphology and ecology and niches has been well studied, how species coexist as diversity is changing, still remains a little unclear. So if phylogenetic constraint is influencing species diversity, then this concept of niche conservatism should be at play. Niche conservatism is the idea that niche related ecological traits are retained over time through a group. And it represents the retention of the number and kind of niches that we see within a group of organisms. So to illustrate this point, I have this little graphic here. And you can see that a similar number and kind of niche, which is represented here in the different colors of circles are retained in two different localities. So this is niche conservatism. And this has mostly been studied in species levels, but there is also evidence of it at higher taxonomic levels. Niche packing on a more local scale is also likely influencing how species are coexisting. Niche packing is a measure of niche overlap as well as niche size. And we expect that with increasing numbers of coexisting species, the sizes of the niches are expected to decrease or remain small. And again, this idea is illustrated in the same little graphic here. The different colors of circles are the different niches. And let's just say, hypothetically, this African river has less species. It is less speciose. Therefore, we're seeing a niche packing pattern of the niches being larger, translate to an Asian river that hypothetically, has more species we're seeing a decrease in the niches. So this is the idea of niche packing, how much overlap and what are the sizes of the niches. And we can explore this in relation to changes in species diversity, and this can help us in understanding modern biodiversity and niche use today. So a way to quantify this idea of niche packing is through disparity. Disparity measures variation or diversity in the different morphologies and shapes of a single group. In a geometric morphometrics context, this is basically the spread in morphospace or the sizes of our niches. Again, a little example graphic here, 'cause I'm so visual. I like to show people what I mean. So here, these circles, this would represent a group of organisms, in my case, a niche. The larger the group, the more disparity. And disparity, like I said, just means that there's more variation or diversity in shapes. So this particular group here with the large circle, that would mean that there's more variation in the shapes of the organisms that are within this group, compared to this one, Partial disparity is the contribution of the subgroup to the total disparity of the more inclusive group. So this is another way of using disparity. This definition seems pretty straightforward until we take a step back and think, okay, but what does total disparity actually mean? And there are a couple of different definitions of total disparity. Foote in 1993, describe total disparity as everything that we would see within this dotted line. So within this dotted circle, that would be Foote's description of total disparity. Adams et al. though, they came along in 2019 and they redefined total disparity as the sum of the partial disparities of the subgroups. So in this graphic, it would be, here's one subgroup, one subgroup, and you would add up those partial disparities to then get a total disparity. This definition makes the most sense to me for total disparity because it gets rid of all of this extra space in between the subgroups that's not actually being occupied by a niche. And when I'm thinking of total disparity, that's what I really care about. I wanna know how much space is being occupied by niches in this morphospace. So for the purposes of this project, I use Adams et al. definition of total disparity, summing the partial disparities. So how can we quantify and assess all of these different things that I just talked about? We can use geometric morphometrics. This is a statistically powerful approach for quantifying body shape. And it's based on homologous landmarks. Now, homologous landmarks, if you don't know, are just places on a body with structural or functional significance. So when we're talking about fish, this could be the insertions of the fins around the eye or the mouth. These are then analyzed multivariately like with the principal components analysis so that you can examine the shape and its variation in morphospace, as well as get estimates of disparity. Essentially, it's a fancy way to quantify body shape as a surrogate for niche position. And we can compare different morphospaces from different communities across localities to see how shape is changing or not. So that brings us to our focal group, Cyprinoidei. Cypriniformes is a globally distributed order of freshwater fishes. This includes fish like minnows, carps, loaches and suckers. And I'm focusing on in the suborder Cyprinoidei. This has 13 families spread pretty much across the whole world in North America, Europe, Asia, and Africa. So pretty much everywhere except for South America and Australia. And this group is cool because they have a huge range of temperature tolerances and diets, and they're found pretty much everywhere that has fresh water. So from large rivers to small creeks and streams and springs. If you can think of it, they're probably there, if it's fresh water. And as a result of these wildly different temperature tolerances and diets and habitats, we also see a lot of variation in body shape within this group. Interestingly, there's also large differences in the latitudinal species richness for this group. So as we get closer to the equator, we see an increase in species richness. Now, all of these things make this the ideal group for this study because we do see all this diversity, especially in body shape and we have changes of species richness. So my objectives. My objective was to investigate niche conservatism and niche packing within Cyprinoidei across a lot of their distribution and to examine any effect that the changes in species richness might have on niche conservatism and niche packing. To do this, I used body shape as a surrogate for niche position using geometric morphometrics and I compared representative stream dwelling cyprinoid communities across different localities. And I looked at niche conservatism and niche packing to see if that was changing with species richness. So essentially, did we find the same number, kind, and size of niche and did that change with richness? So my hypotheses, I had two different hypotheses. The first one is that because cyprinoids share an evolutionary history, they will show niche conservatism. In other words, a similar number of niches will be recovered regardless of species richness. The second hypothesis was that niche packing will change with species richness. So I specifically, hypothesized that with increases in species, we would see a decrease in niche disparity, or essentially the niches would get smaller. And I have another hypothetical graphic here for you to illustrate my hypotheses. So this is what I was expecting to see. Three different hypothetical communities of cyprinoids with differing levels of species richness. Now, the niche conservatism aspect that shows up here in the same number and kind or colored circles showing up in every locality. Niche packing is shown here by the spread in morphospace, so how big are these niches, these circles, and are they overlapping. And I expected that with an increase in species. So let's say, for example, this Pearl River locality had more species. I hypothesize that we had see a decrease in niche disparity. So that's shown here with these smaller circles, compared to the other localities. So getting into my methods. [bottle cap scratching] So I got my specimens from two different museums, the American Museum of Natural History and the Florida Museum of Natural History. When I was deciding which specimens to pull, I limited it to those collected on a single day from a single second to fourth order size stream. I did this to eliminate any environmental variability that could play a major role in species richness across these localities. My sample size was limited to just what was available, what was collected on that day, what the museum had for me, but when possible I did use 30 individuals per species. And this is a map just showing my 14 different localities across Africa and Southeast Asia. So once I had my specimens, I took standard lengths using calipers. I assessed the gill rakers when possible, and I pinned the posterior insertions of the fins. This pinning was just for ease of landmarking down the road. Once they were pinned, I then took pictures of them on their left side in a glass squeeze box with a ruler. And this is an example of one of my landmark images looks like. So I used 21 homologous landmarks. And again, a homologous landmark is just a place of structural or functional significance on the body. So I chose the insertions of the fins around the mouth and the eyes. And these 21 landmarks were used, the exact same ones were used across every single fish. So this is just an example of one fish. And once I had all my landmark images, I uploaded them into this program called MorphoJ and I did a generalized Procrustes superimposition. This just eliminates any extraneous information like the size of the image or the orientation of the fish, if it was accidentally a little bit downward pointed, kind of standardizes all of that information. And then after that, I made covariance matrices using those corrected Procrustes coordinates. Then, I could do a PCA, a principal components analysis to analyze the shape variation that's going on. I also got my principal component values, which basically just represent the majority of the variations that would be PC1 and PC2. And you plot those to then produce a 2D morphospace. I also made wireframe graphs like the one pictured here by connecting homologous landmarks to show the average shape of each niche group, as well as the average shape of a whole locality. And then niche determination. So still in MorphoJ, I made classifier variables based on both species and niche. And these niche classifiers were determined a priori based on morphology and literature reviews that I did. So the main information that gave me these was body shape, mouth position, habitat, and diet. [keyboard clanks] And in total, we found eight niches across all of these different sites. So we had a benthopelagic herbivore, benthopelagic meaning existing anywhere from the bottom to the midwater column. Herbivore, meaning it eats mostly plant material. Benthopelagic omnivore, eating plant and animal material. Benthopelagic insectivore, eating mostly insects. Benthopelagic generalists, eating insects and invertebrates. A benthopelagic piscivore, eating mostly fish. Pelagic omnivore, so moving away from the benthopelagic aspect and really just sticking to the midwater column. Omnivore, again, plant and animal material. Pelagic frugivore, eating mostly fruits and seeds. And then finally, a topwater insectivore, so just existing in that top water habitat and eating mostly insects. So again, a total of about eight niches. I then, to get my disparity values, I ran analyses in R using the geomorph package. And to do this, I used my landmark TPS file. So all that shaped data alongside CSV files to be able to incorporate my niche grouping so I could get disparity by niche group. And the disparity was calculated as the outline of all of the data points in that niche group. I also got partial disparities and those were used in graph for each locality. And then, the total disparity was calculated as that sum of each partial disparity per locality. So again, using the Adams et al. definition of total disparity. I then, made a lot of graphs [laughs] to look at the relationship between diversity and disparity. So first I looked at the species richness across localities. Then I broke it down niche by niche to see the prevalence of each niche at all the localities. And then I did linear regressions to examine the relationship between species richness and total disparity across localities, as well as the relationship between partial disparity and richness broken down by niche groupings that were at four or more localities. So getting into the results. We're gonna take a look at a lot of morphospaces today. We're gonna start with the African localities and their corresponding partial disparities. So here is the first African locality in Liberia. Now, this might be a little overwhelming, especially if you are unfamiliar with morphospaces in these concepts. So no worries, I'm gonna walk us all through it step by step. So to the left is a morphospace. Each individual dot that you're seeing are the summed landmark data for one individual fish. So this dot's a fish, that dot's a fish, they're all fish. And they're color coordinated by species. And they have these 95% confidence ellipses by group, by niche group, sorry. And then, the important part is that you can see the average shape for each niche group in these wireframe graphs. So I'll walk you through that in a little bit more detail. If you could focus in on this benthopelagic herbivore wireframe graph, the light blue here, that shape is the average shape for every fish in this morphospace. The dark blue line is the average shape for the fish in this particular niche group of benthopelagic herbivore. So this is really cool because we can see how the shape is changing niche by niche or not from the average shape. So that's kind of how to interpret what we're seeing there. And then to the right, we have all of our partial disparities broken down by niche. And again, disparity just means, if you have a larger disparity, you have more diversity or variability in the shapes of a group. So for this particular locality in Liberia, we have three different niches, benthopelagic omnivore, pelagic omnivore, and benthopelagic herbivore. Pelagic omnivore had the largest disparity or variation in shape, benthopelagic herbivore had the second most, and the omnivore had the last. Moving on to our next locality in the Democratic Republic of the Congo, we have another three niche groupings, benthopelagic piscivore, pelagic omnivore, and benthopelagic herbivore with the herbivore having the largest partial disparity and the pelagic omnivore having the smallest. Now, something I want you to notice in this particular morphospace, it's very clear that there are two distinct groups within the benthopelagic herbivore niche grouping. Now my hypothesis is that given more accurate niche data, this would actually be two separate niche groupings. So this is something that perhaps another study could look into. The locality in Ethiopia just had two niche groupings of benthopelagic omnivore and benthopelagic herbivore with the herbivore having a larger disparity. The first South African locality had two species, therefore two niche groupings of a benthopelagic omnivore and a benthopelagic insectivore. These disparities were very similar, but the omnivore had a slightly larger disparity. Second South African locality. Again, two niche groupings, benthopelagic generalist, benthopelagic omnivore. And again, very similar disparities with the generalist having slightly larger. And something to note here, this is our first case of niche overlap, which would be part of a niche packing pattern. However, something to notice is that the actual data points are not overlapping at all. It's just the 95% confidence ellipses. So there is some overlap. However, these niche groupings are still fairly distinct. Moving into the Southeast Asian morphospaces and their respective partial disparities. So the first Thailand locality had two niche groupings of benthopelagic omnivore and pelagic omnivore with the omnivore having, sorry, the benthopelagic omnivore having a much larger partial disparity. The second Thailand locality had two niche groupings again, the topwater insectivore and the benthopelagic omnivore, larger disparity here. And the third Thailand locality had three niche groupings, benthopelagic generalist, benthopelagic omnivore, benthopelagic insectivore, with the omnivore having the largest disparity and the generalist having the smallest disparity. The first Malaysian locality had three different niche groupings, benthopelagic herbivore, pelagic frugivore, and benthopelagic omnivore, with the benthopelagic herbivore having the largest partial disparity and the frugivore having the smallest. The second Malaysian locality had three different niche groupings, benthopelagic generalist, an omnivore and an herbivore, with the herbivore having the largest and the omnivore having the smallest partial disparities. The first Indonesian locality had two niche groupings, benthopelagic generalist and an herbivore, with the herbivore having a much larger disparity. The second Indonesian locality had four niche groupings this time, very exciting. So we had a benthopelagic insectivore, benthopelagic omnivore, a pelagic omnivore, and a benthopelagic herbivore, with the insectivore having the largest partial disparity and the benthopelagic omnivore having the smallest. Again, take note in this particular morphospace, we have a case of overlap between the pelagic omnivore and benthopelagic herbivore. This is a larger case of overlap than that first one we saw. So keep that in mind. The Bangladesh locality had three niche groupings, benthopelagic insectivore, benthopelagic omnivore and topwater insectivore, with the topwater insectivore having the largest disparity, insectivore having the smallest. The Vietnam locality had another three niche groupings, benthopelagic generalist, benthopelagic insectivore and benthopelagic herbivore, with the generalist having the largest disparity and the insectivore having the smallest. Okay. Moving away from all those morphospaces. We're gonna look at some graphs now. So first, looking at species richness over all of the localities. We had the smallest species richness at two species for a couple of these localities and the largest was 13 in the Democratic Republic of the Congo. [keyboard clanks] Now, when we break it down by niche, we see that the benthopelagic piscivore and the pelagic frugivore niches were only present at one locality each. Whereas the benthopelagic omnivore was present at 12 of the 14 localities. Now, looking at the partial and total disparity values, the benthopelagic insectivore at the first South African locality had the smallest partial disparity while the topwater insectivore in the Bangladesh locality had the largest. By average partial disparity, the pelagic frugivore had the smallest and the topwater insectivore had the largest. And finally, when we look at total disparity at each locality, the first South African locality had the smallest total disparity. So the least amount of space being taken up by niches and Bangladesh had the largest. So looking at our linear regressions now, we'll explore the relationship between disparity and species richness. So total disparity went up as species richness went up for most of these localities. So this was a positive relationship and we did have a statistically significant P value of 0.016. When we break it down niche by niche, we don't quite see that relationship. Or yes, we, sorry. We do see that relationship. So these niches, sorry, the next couple of graphs are for the niches that were present at four or more localities, the first one being the benthopelagic herbivore. So here, we have a neutral relationship between species richness and partial disparity, but it is not statistically significant with a P value of .75. Benthopelagic omnivore, now we do see a slight positive relationship between richness and disparity, but again, not statistically significant. Benthopelagic insectivore, a stronger positive relationship here, but again, not statistically significant. And the benthopelagic generalist, a more neutral relationship or no relationship between richness and disparity, not statistically significant. And finally, the pelagic omnivore, which was our only negative relationship. So this is the only one that actually upholds my hypothesis of niche packing, but again, P value of 0.12, not statistically significant. And this was the only negative relationship. This is the exception. So moving into the discussion, we'll start with niche conservatism. My first hypothesis, if you remember back to the beginning, was that due to their shared evolutionary history, cyprinoid fishes would show niche conservatism with that similar number of niches being retained across the localities. And we did find that, we found a similar number of niches about two to four being recovered across sites. And with more species present, there were usually more niches present, which makes sense. However, there were never more than four niches, even when species richness was at its highest. So it does seem to cap out at around four niches when you have more species. So due to this evidence, I failed to reject this first hypothesis. It does seem that Cyprinoidei is exhibiting niche conservatism with about two to four niches being recovered. [keyboard clanks] So like I said, there were similar number of niches, but the types did vary with a few being found at many of the sites. So the benthopelagic omnivore was found at 12 of the 14 sites, almost all of them. Now, omnivores can be classified as a more general diet. And there is this global pattern of increasing omnivory in fish with a decreasing latitude. And given the tropical location of many of the sites that I used, the fact that we got so many omnivores does uphold this pattern. Now, the rarer niches that were found at one or two sites each, like the topwater insectivore, benthopelagic piscivore and pelagic frugivore, these all have much more specialized diets compared to the more general diet of an omnivore. And so the rarity of these niches could be tied to the availability of the resources that they need. Now, you may have heard me say the word benthopelagic about a hundred times so far. That's because there's a commonality among these localities in the benthopelagic aspect of the niche. Five of the eight total niches had a benthopelagic aspect and there was at least one benthopelagic species across all 14 sites. So this is a fairly general term. It just means you're living and feeding near the bottom, as well as into the middle water column. So these fish are opportunistic feeders that are forging kind of at all depths. And this is showing us that there could be a more generalized trend of seeing a more general species in the benthopelagic area. Now, these trends in niche conservatism that we're seeing could potentially be the result of conversion evolution. Kind of going back to the concepts we talked about at the beginning, similar selection pressures are resulting in similar morphologies and niches and selection has driven cyprinoids to converge on two to four niches of eight different kinds at a locality. And again, we saw minimal overlap of niches indicating that they are distinct from one another. Moving into niche packing. So my second hypothesis was that with an increase in species richness, the disparity of the niches would decrease. And we found the opposite to be true. We found that increasing total disparity was found with an increasing species richness. And even when we broke it down, niche by niche, a similar relationship or no relationship at all was recovered. So due to this, I do reject this second hypothesis. [keyboard clanks] And now the reasoning behind that hypothesis was density dependence or competition as a driver for diversity and coexistence. So the idea is that if you increase competing species, species will either go extinct or become more specialized in their niches, thereby allowing organisms to coexist. And if you more specialized, [computer ringing] you'll probably have a smaller disparity. So that was the reasoning behind that hypothesis. So why did we find the complete opposite pattern? Well, I have an alternative theory and it's the benefits of being a generalist, which I'll discuss now. [bottle cap scratching] [bottle thuds] So with increasing competition from increasing species richness, it may be more beneficial to be a generalist species. You're able to take advantage of a greater breadth of habitats and prey, especially given seasonal changes and really heterogeneous environments. Many of the sites that I used in this study in Southeast Asia experience flooding and monsoon rainfall. If you're a generalist, you would be able to capitalize on these resources as they become available with your changing environment. So again, a larger disparity means you have a larger diversity of morphologies in a niche, more generalized, and you can capitalize on a wider array of habitats and prey. So when you're occupying a more general niche with the larger diversity of morphologies, you can cope with this increased species richness. And that might account for the increase in disparity that we saw. And again, when we broke it down niche by niche, there was a similar trend or no trend found at all. So we had two positive relationships, two neutral relationships and one negative relationship, but none of them were statistically significant. The two positive relationships that we saw, could be explained by that generalist theory that I just discussed. And then, the two neutral and one negative relationship might be explained by competition driving coexistence, which was my original hypothesis. However, the evidence is still not strong enough, especially without that statistical significance. And so I still do reject that second hypothesis. Now, I would be remiss if I didn't bring this up. There are, of course, species outside of Cyprinoidei existing in their communities that would likely also play a role in these niche packing patterns. Due to their prevalence and co-occurrence, Characins in Africa and Osphronemids and Channids in Southeast Asia are very likely influencing niche packing patterns of Cyprinoidei. And so if we did include them, it would, of course, influence the disparity of niches that we're seeing occupied by cyprinoids, but we did not include them for this study and therefore there are effect is largely removed. So conclusions, to wrap everything up, was niche and conservatism upheld regardless of richness and composition? Yes, there were about two to four groupings per site, capping out at about four niche groupings. Did disparity decrease with that increase in species richness? Nope. We saw the opposite. We saw an increase in disparity with an increase in species richness. And again, that could potentially be explained by the benefits of being a generalist, which I discussed previously. So more work is needed on a lot of different fronts to bring more clarity and context to this, the results of this project. First up, resolution of phylogenetic relationships. So unfortunately, a lot of the species that I used in this study have still unresolved phylogenetic relationships. So if there was a more clear relationships for these species, I think it would give a lot of more historical context for the patterns that we see. Inclusion of species outside of Cyprinoidei, so like those Characins and Channids that I was talking about a few slides ago, this would give an interesting and more intact community picture for niche packing patterns. And I think that would be a fun study to do. And complete life history data is really lacking for a lot of these species that I used in this study. So stable isotope or trophic studies are needed for many of these species to get more correct diet information and more detailed information on habitat preferences is needed as well. This would really help with a more accurate defining of niche groups within these morphospaces. And finally, more studies exploring this idea of the benefits of being a generalist, especially in this context of niche packing. So deep breath, everyone, we're now gonna totally switch gears and I'm gonna introduce you to my dissertation. So what I'm currently working on. My first chapter is the identification of larval Pacific sand lance, Ammodytes personatus and Arctic sand lance, Ammodytes hexapterus, using CO1 and morphological characteristics. So starting again with some background, many of you may already be really familiar with this, but you're gonna get it anyways, just so that everybody is on the same page. So forage fish include lots of fish like anchovies, menhaden, herring, and what I really care about, sand lances. These are highly abundant, small schooling pelagic planktivores, and they're crucial parts of the marine food web because many sea birds, marine mammals, and other commercially important fish are depending on them for food. Sand lances are especially important because they really help in the transfer of energy between producers and upper trophic level species, as well as between pelagic and benthic regions. They're very abundant, though that abundance does vary year to year. And of course, that is then going to affect all the populations of organisms that are feeding on them. One of my favorite and most interesting parts about sand lance is that they burrow. These fish burrow headfirst into the substrate, and then they're able to kind of swim or undulate around in the substrate to get positioned. Now, from the reading that I've done, it seems that most of our knowledge comes from Ammodytes personatus. The substrate that they prefer is coarse sands, especially as a burrowing species, substrate is very important to them. Depth, they're found in nearshore habitats about 60 to 80 meters. So they have been found deeper. Their diet seems to mostly consist of copepods, a few amphipods and polychaete worms. And in terms of temporal patterns, they show crepuscular foraging, which means that they forage at dusk and at dawn, and they burrow in between to escape predation. They also show a longer period of burrowing in the winter. So they have a winter dormancy period, and then they come out and they forage more often in the spring, summer, and fall. Sand lances have a complicated taxonomic history. There's currently six recognized species in the genus Ammodytes, with a recent separation of the Arctic sand lance, which is Ammodytes hexapterus and the Pacific sand lance, which is Ammodytes personatus. Now, the larvae of these two species are really similar morphologically and they have overlapping ranges, which provides some challenges, but also some really interesting questions to ask. So this map comes from Orr et al. 2015, and it's a nice way to show you the two species ranges and where they're overlapping. So the Pacific sand lance, Ammodytes personatus, are these filled in circles here, you can see it has more of a Southern range coming up into the Gulf of Alaska through into the Eastern Bering Sea. Ammodytes hexapterus, the Artic sand lance is the open squares, a more northerly distribution up in the Chukchi Sea, coming down south into the Eastern Bering Sea. So again, the Eastern Bering Sea is that area of overlap. And this is really the basis for my questions because the larvae looks so similar and they have this area of overlap, we really need to be able to visually ID and tell these larvae apart so that when they're caught, especially if they're caught where they're overlapping, we know what we're looking at and we're able to ask more questions about their early life history and ecology. So the objectives for this first chapter, I'm wanting to identify the larval Ammodytes personatus and Ammodytes hexapterus doing two things, using molecular techniques to confirm the idea of the larvae from across the distributions and also to examine the morphological diagnostic characters so that we can visually ID the larvae. So the next few slides are going to be from a pilot study that I did over the last couple of months, just to kind of see what's going on and what direction we need to go with this. So this map is from the EcoFOCI Cruise Catalog. And the samples that I got are from the AFSC Archive and the data like catch location and everything, was also from the Archive and the EcoFOCI databases. For this first pilot study, I did a total of 47 samples, 15 from the Gulf of Alaska here. These 15 are presumed to be all personatus. 11 from up in the Arctic in the Chukchi Sea. These 11 are presumed to be hexapterus. And then finally, 21 specimens from their area of overlap in the Bering Sea, and these were assumed to be a mix of both species. So starting with the molecular work, I dissected out the right eyeballs of the larvae when available, if the right eyeball was already gone, then, of course, I took the left eyeball. And then from there, DNA was extracted using the QIAamp micro DNA kit. And then I did PCR on the fixed nucleotide differences in the mitochondrial DNA genome at the CO1 gene. I did gel electrophoresis, sent everything off for sequencing. And then when I got my sequences back, I trimmed and aligned them and made consensus files in this program, Geneious Prime. I then got other sequences of Ammodytes from the NCBI GenBank, and I made a tree. So here's the main results for this molecular work. And I know that you probably can't actually read all of the tip labels for this tree. That's okay. Don't worry. I'll tell you what you really need to know. So, first of all, of the 47 samples, we did get three samples back that were not Ammodytes. So we got a Lumpenella, an Anisarchus and Leuroglossus. This simply came down to mis-ID at the beginning, and they were mis-ID'd as Ammodytes when they were, in fact, not. These twos clades here are all of the extra Ammodytes specimens that I pulled from GenBank to make a more complete tree. So this is like Ammodytes japonicus, americanus, heian, dubius, all of those species. Then, the important part for us, this clade is all Ammodytes personatus, And all of the Gulf of Alaska samples did come back as personatus. This is good. This is what we expected. All of the Arctic samples came back as Ammodytes hexapterus, this clade here. Again, this is good and what we expected. Now, the more interesting part were the 21 Bering Sea samples. So we had a total of 21 in their area of overlap, of those 20 came back as Ammodytes personatus. So almost all of them came out in this clade as personatus and only one Bering Sea sample came back as hexapterus. So I zoomed in here on the tree for you so you can see our one lonely Bering Sea hexapterus. And this is interesting for a couple of reasons. So this map again pulled from the EcoFOCI Cruise Catalog. And I have zoomed in for you here with this red circle on Station 45. This is where that one Ammodytes hexapterus was caught in the Bering Sea. This is interesting because it's much farther south in their area of overlap than we expected to be catching hexapterus. We really thought that hexapterus would be caught farther up north kind of by Nunivak Island. So the fact that the one that we did get was so far south is quite interesting. It also poses the question, is this a fluke that there's only one out of 21 that came back as hexapterus? Or is that really the proportions of the two species of larvae that we're seeing? So moving into the morphology part, once I got the molecular IDs, I examined the larvae for morphological diagnostic characteristics so that we could visually ID them. Now, so far, what I think that I'm seeing is that pigment is the primary distinguishing characteristic. So specifically, seeing dorsal and cranial pigment, as well as ventral post-anal pigment as the differences. And the overall trend or pattern that I think I'm seeing is less pigment in Ammodytes personatus and more pigment in Ammodytes hexapterus, also at smaller sizes. But this comes with the caveat because when you compare the molecular ID with the hypothesized IDs that I did just solely based on morphology, it only comes out about 70% correct. So there's clearly a lot of refinement that needs to happen with these distinguishing characteristics and, you know, quantification in all of that, because ideally we would have closer to 100% with our visual ID skills. But just to give you a little taste of what I'm seeing and the patterns that I think that I'm seeing. So I have an Ammodytes personatus at about 10 millimeters and then a similarly size hexapterus. When we look dorsally, you can see that there's barely any pigment at all on the personatus compared hexapterus of a similar size. We do have pigment going on, especially on the cranial area. Eventually, the personatus does have a bit more pigment, but still when you compare post-anal melanophores here, the hexapterus does have more and they seem to be darker. And a similar pattern comes up at a larger size. Again, dorsal view of personatus, very few pigment patterns going on here, especially compared to the hexapterus of a similar size, a lot more pigment. And eventually, a similar pattern with the hexapterus having more and darker melanophores here, post-anally. So my next steps for this first chapter are to continue testing and refining these diagnostic characters on a second and larger set of individuals. Specifically, I wanna get a lot more samples from the Bering Sea to look into the question, how many hexapterus are there actually in the area of overlap? And how far south are we finding them? And if I'm able to successfully ID visually these larvae and answer these questions about hexapterus, this would open up a lot of different research questions on their ecology and early life history, especially where they're overlapping in the Bering Sea. Specifically, questions that I'd like to ask in my second chapter are, how are these larvae of the two species coexisting? Are they niche partitioning? If they are niche partitioning, what aspect of their niche are they partitioning? So hopefully stay tuned. Maybe I'll be able to have some data to present on that soon. And with that, I'll start with acknowledgments for my Masters. I'd like to thank the collections, the curators and staff at the American Museum of Natural History and the Florida Museum of Natural History for letting me come and use their specimens. Funding from the ASIH Raney Award and the BGSO Travel Grant that allowed me to travel to these museums. My Master's advisor, Dr. Kyle Piller for all of his help and support. The Piller Lab, and especially my former lab mate, Sarah Ward, pictured here. She actually traveled with me to the museums and helped me collect my data and take all my pictures and I could not have done it without her. Acknowledgments for my PhD. Of course, I'd like to thank my advisors, Luke and Ali, they're amazing. And they have helped and supported me so much along the way. And a ginormous huge thank you to Melanie. Thank you, thank you, Melanie. Melanie helped to train me in DNA extraction. She also sat down and talked me through, you know, where to find the species or sorry, the specimens, her ideas on the area of overlap, all sorts of stuff. And she also let me use some of her Arctic samples. So big thank you. Thank you to the AFSC Archive for the use of the specimens and data along with EcoFOCI, for their help in the data that I get. And the University of Washington Burke Museum Fish Collection, and the Tornabene Lab for all of their support. And with that'll take any questions.

[Deana] All right. Great job. I'm just gonna clap for everybody since everyone is muted.

Thanks.

That was a great talk. Thank you so much, Laurel. And yes, let's open the floor up to any questions.

[Laurel] Here, I think I'll stop sharing for now so that I can see the chat, but if we need to look at any slides I can share again.

[Deana] Perfect. [computer dings] [bottle thuds] Give everyone a minute to warm up. [laughs] All right. Any questions from anyone? All right, here we go. Pascal asks, "Have you entertained any thoughts on venturing outside of PCA to do the clustering classification like using deep learning?" Oh, that's a great question.

[Laurel] Yeah, that's fascinating. I don't know if I've actually, if I know exactly what you're meaning by deep learning. I'll have to look into that. From all like the research that I'd done leading up to that Master's project, the PCA was really the only method that I had seen consistently used. So I'll definitely need to look into that. That sounds really interesting.

[Deana] Pascal followed up and said K-means.

[Laurel] Yeah. I'm not sure that I've heard of this. [computer dings]

[Deana] We love great ideas.

[Laurel] Yeah. I like the sound of deep learning too. That sounds interesting.

[Deana] Yeah. Pascal says it's okay. I haven't seen PCA used for classification in a while, so I'm interested in your dissertation.

[Laurel] Interesting. Okay, thank you, Pascal. Yeah, I'm definitely gonna look into these methods you were talking about.

[Deana] All right. Here's a comment from Ally. "Great talk, Laurel. For the sand lance work, do you have any hypothesis on what may be driving speciation between the two sand lance species of interest?"

[Laurel] Yeah. Thank you, Ally. That's a great question. I do have a little hypothesis just based on a couple of little trends that I think that I'm seeing, but not actually based in any real data, so we'll see. I'm really interested in looking at the burrowing mechanisms and the morphology behind that. So the sand lance have they, over time, they evolve or develop, sorry, these projections coming off of the dentary and that's the first thing that hits the substrate when they're burrowing head first. And from what I've been looking at in the developmental series, it seems like hexapterus might be evolve or not evolving, sorry. I keep using that word, developing those earlier than personatus. So I have this little hypothesis that maybe there's a difference in the grain size or a slight difference in the grain size between these two species and maybe they're burrowing slightly differently, something like that. So that's kind of a hypothesis that I'm might wanna look at in the future. Does that answer your question, Ally? Cool.

[Deana] Yes. Thank you. Very neat. [Laurel laughs] Yeah. I learned a lot about sand lance. [laughs]

[Laurel] Good.

[Deana] Ryan asks, "if you can explain again, why the plateau at two to four niche groups?"

Yeah.

Is it impacted by the environment at all? Or what other environments give rise to higher numbers?

[Laurel] Yeah, that's a fantastic question. I think that it's because of the size of stream that I was specifically limiting myself to. So like I said, I was looking at like second or third order streams, which is gonna limit species richness and therefore limiting niche groupings, and also how many habitats can be in a certain size stream. So if we're looking at a larger river, I do think that that's gonna cap out more, maybe like 10 or 12 niches, something much bigger. So I do think it scales with the environment and communities that you're looking at, if that makes sense. Thank you for the question.

[Deana] Yeah. Great answer. All right. We might have time for one more question. We're coming up to the hour. I wanna thank everyone for their questions and Laurel for your presentation. Ryan says he was surprised to hear about fruit eating fish.

[Laurel] Yeah, those are so cool. You should definitely look into those more. I love the frugivores. [Deana laughs] Well, thank you so much, everyone.

All right.

I really enjoyed giving a talk. Thank you so much for coming.