COMPUTER ANNOUNCEMENT: This conference will now be recorded. HEATHER TABISOLA: Good morning, everybody. Welcome to another EcoFOCI seminar series. I am Heather Tabisola. I'm co-lead of the series. And this week, you get to meet Deana. Deana, say hi. So Deana will be leading the series with me for the at least foreseeable future. [LAUGHS] We are part of NOAA's EcoFOCI biannual seminar series, which is focused on the ecosystems of the North Pacific Ocean, the Bering Sea, and the US Arctic to improve our 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 ecosystems. And you can visit our page, which is ecofoci.noaa.gov, for more information on the program. We once again thank you for joining us. It's another virtual season. We know everybody's getting tired of it, but we sincerely thank you for being here today and listening to our speaker. You can find our lineup via the OneNOAA Science Seminar Series and also on the NOAA PMEL calendar of events. This is our second of four this season. And then we will be back in the spring. So we are here at 10:00 AM Pacific time on Wednesdays for today and the following two weeks. And if you happen to miss a seminar, you can catch them on PMEL's YouTube page. It does take a few weeks, traditionally post season, to get these up on the web. But we will have them. So if you have an immediate request, please do reach out to the speaker directly. Double check your microphones. Make sure that you are muted, that you are not using video. And then during the talk, please feel free to type your questions into the chat at any time. Deana and I will both be monitoring to those. And we'll address the questions at the end. So today, I am pleased to introduce Dr. Jennifer Bigman, who joined our program on the AFSC side in July of 2021. She is a postdoctoral researcher. She completed her PhD at Simon Fraser University where she worked on metabolic ecology; the importance of respiratory surface area for scaling metabolic rates across species; and links between physiology, morphology, and life history in fishes, especially sharks. Dr. Bigman is working with Dr. Lauren Rogers and Dr. Ben Laurel, along with a larger team of collaborators to study how spawning habitat for Pacific cod is changing in the Bering Sea. Today, Jennifer will be sharing her talk on the connection among physiology, ecology, and life histories from a macro ecological perspective. And with that, let's begin. JENNIFER BIGMAN: Thank you. Here we go. So thank you all for attending my talk today. I'm going to be talking mostly about my PhD work, which, like Heather said, I recently finished in June at Simon Fraser University. And when I was there, I worked with Nick Dulvy and Nick Wegner, who is a research fisheries biologist at NOAA Southwest Fisheries Science Center in La Jolla. And so even though most of my talk will be on my PhD, I'm going to talk a little at the end about what I'm working on at NOAA. So to start, I just wanted to take a second and define macroecology, which is really my background and training and generally frames how I think. So macroecology is a subfield in ecology. And it's concerned with understanding the generality of relationships among organisms and their environments. And this is typically at large spatial scales. And historically, the goal was to understand broad patterns of abundance, distribution, and diversity. And so originally, it was just really a new name for the field of biogeography. But as time goes on, this field has really expanded. And many have applied macroecological framing to other questions including understanding and predicting extinction risk and species responses to climate change. So a big part of macroecology revolves around ecological theory and specifically developing and testing theory that attempts to explain observed widespread patterns. And this is really a big part of my work specifically because I test theories that try to predict how environmental oxygen and temperature will scale out to affect species through their organismal physiological traits. And so oxygen in particular is really central to the story because it fuels most of life on Earth. And this is both in aquatic and terrestrial systems. And of course, this is largely because the majority of organisms, particularly the invertebrates, respire aerobically, which means that they need oxygen to fuel metabolic processes. And these metabolic processes broadly function to transform resources from the environment into available energy. And I can't really define metabolism without showing you this short clip from Parks and Recreation where Andy Dwyer explains metabolism really better than I ever could. See if it'll play. [VIDEO PLAYBACK] - Did you know that the food you eat becomes energy? Yeah. Boom. That's spaghetti. Nachos. That's a cookie. [END PLAYBACK] JENNIFER BIGMAN: And this energy that's produced for metabolism is then allocated to life sustaining processes such as growth, survival, and reproduction. And metabolic rate is the rate at which this transformation of energy occurs. And it's so intertwined with oxygen that it's typically measured as an organism's oxygen consumption over unit time, which is usually done by respirometry in a laboratory setting. But acquiring the oxygen needed to fuel metabolism is dependent on many organs, systems, and pathways, which results in a host of other traits that are related to oxygen consumption. For example, the respiratory organs-- so this would either be the gills for water breathing organisms or lungs for air breathing organisms-- are key because the oxygen needed to fuel metabolism is diffused over the surface area of the lungs or the gills. And oxygen in terms of both metabolic rate and respiratory surface area and more broadly the balance between oxygen supply and demand is invoked to explain several macroecological or widespread patterns. And so, for example, the temperature size rule is the reduction in the maximum body size and faster growth of ectotherms with increasing temperature. And this was originally focused on lab reared organisms with similar patterns in the wild called either James' rule if you're talking about this pattern within species and then Bergmann's rule if you're talking about this pattern across species. But generally, the pattern of faster growth to a smaller maximum size in warmer temperatures in the lab and in the wild are today generally all referred to as the temperature-size rule. And so this pattern has been suggested to be mechanistically related to oxygen for those species that are aquatic. And this is because warmer waters generally have lower oxygen availability. So the hypothesis here is that there is oxygen limitation or a mismatch in oxygen supply and demand. And so more recently, work has also shown that the geographic distribution of marine species is not only correlated with temperature as we're all probably very familiar with, but that oxygen plays a really important role as well. And so on the other hand, we have several macroecological theories that also center on the importance of metabolic rate, respiratory surface area, and the balance of oxygen supply and demand. So for example, the metabolic theory of ecology, it aims to leverage the body mass and temperature dependence of metabolic rate by using the relationship of organism metabolic rate, body mass, and temperature across species-- so really broad here ranging from bacteria to whales-- to predict patterns and processes across all levels of biological organization such as development and mortality rate, population growth, and even ecosystem level processes such as biomass cycling. But in reality, the capacity of this theory to actually predict these phenomena, especially those that extend beyond the individual or species level processes such as ecosystem cycling or biomass cycling, have not really been fully realized. And in part, this is because even after body mass and temperature are accounted for, a considerable amount of variation in metabolic rate remains to be explained. And this suggests that other traits may help shape this relationship, perhaps those such as respiratory surface area. And the intimate relationship between metabolic rate and respiratory surface area forms the basis of another macroecological theory which is called the gill oxygen limitation theory. And this theory argues that the surface area of gills of aquatic water breathing organisms limits aerobic metabolic rate and ultimately growth in other processes that rely on the energy produced by aerobic metabolism. And the central tenet of this theory is that the surface area of the gills, which are roughly a two dimensional surface, cannot grow as fast as the body they have to supply with oxygen, which is roughly a three dimensional volume. So there's a mismatch in geometry. And this means that the slope of the relationship between whole organism gill surface area and body mass within a species-- so we're here we're talking ontogenetically-- would be geometrically constrained to be of value of less than 1. And so if we look at this from a mass specific perspective-- so now the y-axis is relative gill surface area, which is just the ratio of gill surface area to body mass or the gill surface area per unit of body mass-- this ratio decreases throughout an organism's lifetime and as espoused in the gill oxygen limitation theory, would result in a decline in the relative oxygen supply as an organism increases in size. And eventually, this theory suggests or indicates that the oxygen supply from the gills would only be enough to fuel maintenance metabolism, which under this theory is defined as the minimum oxygen supply required to maintain a body mass but not grow. And thus, once the gills can only fuel this maintenance metabolism, which is where the blue and yellow line intersect, the organism will stop growing, and its maximum size will be reached. And interest in these broad macroecological theories and the role of oxygen in ecology and evolution has experienced a recent resurgence because of the implications of a changing climate, namely higher temperatures and lower oxygen availability. So for example, the metabolic theory of ecology would predict that there would be an increase in temperature, which would result in an increase in metabolic rate, and that these changes would scale up to affect other patterns and processes across all the levels of biological organization such as those that I mentioned earlier. So likewise, the gill oxygen limitation theory makes explicit predictions about how increasing temperature will affect fishes and other water breathing aquatic organisms. So if we go back to this conceptual figure that I showed you in the introduction where the relative oxygen supply declines with size, the gill oxygen limitation theory would predict that an increase in temperature would result in an increase in maintenance metabolism and ultimately a smaller maximum size and presumably a reduction in the available energy for other processes in addition to growth. And this theory suggests that oxygen limitation, which is this idea that the supply of oxygen from the gills cannot keep up pace with oxygen demand, has been used to explain several other key patterns and processes in fishes such as the timing of maturation, spawning dynamics and behavior, as well as geographic distributions and even patterns of habitat use. But this theory and really more broadly the role that oxygen plays in shaping ecology and evolution particularly in light of climate change are hotly debated in the literature. And moreover, most work in this area, which examines the links among oxygen, temperature, physiology, and ecology, have historically been experimental in nature and have focused on the observations and responses of single species in laboratory settings. And so, of course, this work is most certainly necessary. But it generates pieces of a much larger puzzle that can be put together by meta analysis and modeling to understand the generality of relationships. And so by coupling field collections, laboratory dissections, and meta analysis and modeling, I sought to understand the generality of the relationships among traits related to oxygen acquisition and use, ecology, and life histories in fishes and other vertebrates for my PhD. And so under this overall question, I have three main questions that I'm going to talk about today that roughly correspond to links in this conceptual figure. So the first question corresponds to this link right here in the middle. And it asks whether respiratory surface area helps us understand the immense variation in metabolic rate with body size across the vertebrate tree of life. And then the second question corresponds to this link between ecology and respiratory surface area and looks at how ecological factors relate to gill surface area in sharks. And then finally, the third question asks how gill surface area relates to life histories across fishes. And so I'm going to walk through each of these main questions separately. So for this first question, the basis for it, the first question that examines whether respiratory surface area can explain variation in metabolic rate, the basis of this question really stems from the foundational relationship of the metabolic theory of ecology, which is that among metabolic rate body mass and temperature. And as I shared with you in the introduction, the metabolic theory of ecology proposes that at a first approximation, body mass and temperature explain much of the variation in metabolic rate across species. And this is from bacteria to whales. But a considerable amount of variation in metabolic rate still exists even after we account for the variation that body mass can explain and the variation that temperature can explain. So specifically, metabolic rate for organisms of the same body mass varies up to five orders of magnitude. And that's even after you account for body mass and temperature. And given the intimate relationship between metabolic rate and respiratory surface area, I wondered whether respiratory surface area could explain this variation in metabolic rate. And so to answer this question, I collated estimates of metabolic rates and respiratory surface areas and their associated body masses-- so their measurement body masses-- and temperature for as many vertebrate species as I could find. And so this resulted in a data set of over 100 vertebrate species with representatives from all major lineages for which estimates of both metabolic rate and respiratory surface area existed. But in compiling this data set, I found that metabolic rates and respiratory surface areas have rarely been measured at the same body masses in the same species. And so in this figure, each data point is the absolute percentage difference in body mass between metabolic rate and respiratory surface area for each species in the data set. And any species data point that is not red or is above that red line indicates that there's a large difference in the body mass for which metabolic rate and respiratory surface area were measured, which is the case for most species, as you can see. And this difference in body mass really complicates the comparison of metabolic rate and respiratory surface area across species because both are size dependent traits, which means that they vary with body size. So we know that a mean value wouldn't really capture the variability in each trait with size within species. And so to solve the size mismatching of traits, I and my colleagues developed a phylogenetic Bayesian multilevel modeling framework that enabled a vertebrate wide analysis of the size dependent phenomena while also accounting for other covariates. And so just to give you the nuts and bolts of this model, the first level estimated the relationship between respiratory surface area and the measurement body mass associated with respiratory surface area. The next level estimated the residual respiratory surface area, which is just the difference between each species observed respiratory surface area value and the fitted value predicted from the model. And then the last level use thesed residual respiratory surface area values as data in a model that examined the relationship between metabolic rate, body mass temperature, and any other covariates such as thermoregulatory strategy, which is whether a species is an endotherm or an ectotherm. And I'll explain in a moment why that is so important. And I also included a random effect of phylogeny to account for the shared evolutionary history among species because when you're working across so many species, they're not independent from each other because they share varying parts of their evolutionary trajectory. So some species share more than others. And so a trait value of some species might be more similar because they're related and not because of the ecology. And so the neat thing about this model is that it propagates the uncertainty and the residual respiratory surface area across all levels of the model because each iteration happens in succession. And so I used this modeling framework in an information theoretic approach to assess whether respiratory surface area explained variation in metabolic rate across vertebrates while accounting for the other important covariates. And I found that even after we accounted for body mass, temperature, and thermoregulatory strategy as well as evolutionary history, species with higher metabolic rates, which is on the y-axis in this figure, for their body mass, which is on the x-axis, had greater respiratory surface areas. And so in this figure, species with an orange have higher than expected respiratory surface area for their body mass. And species in purple have lower than expected respiratory surface areas for their body mass. And the fact that species with higher metabolic rates had greater respiratory surface areas was exemplified when you compare organisms of the same body mass. So for example, the endothermic kowari rat had about the same mean body mass as this fish, which is the ecotothermic white sucker. But the kowari rat had 32 times greater residual respiratory surface area and 16 times greater metabolic rate compared to the white sucker even after I counted for differences in temperature as well as the fact that the endothermic kowari rat was an endotherm. And so when we compare the relative importance of these predictors of metabolic rate in terms of standardized effect sizes, we see that compared to temperature, residual respiratory surface area explains up to twice as much variation in metabolic rate. And you can see that because the mean effect size of the residual respiratory surface area is twice that of temperature. And so this is a pretty big finding because for over 20 years, we've thought that most of the variation in metabolic rate is due to body mass and temperature. But before I got too excited about this result, I first needed to make sure that the importance of respiratory surface area wasn't simply because it was just the same as thermoregulatory strategy. And this is because we know that endotherms have higher metabolic rates for their body size compared to ectotherms. And this is because endotherms retain metabolically produced heat to regulate their body temperature within a relatively narrow thermal range. And so an endotherm's already going to have a higher metabolic rate for their size than an ectotherm. And so to ensure that the importance of respiratory surface area wasn't simply because it was directly related to thermoregulatory strategy, I examined the scaling between metabolic rate and body mass, which is going to be on the top, and respiratory surface area and body mass, which is on the bottom for endotherms in red and ectotherms in blue. And we would expect that there would be a match in scaling between metabolic rate and respiratory surface area regardless of the thermoregulatory strategy if respiratory surface area and thermoregulatory strategy were explaining the same proportion of variation in metabolic rate. But for endotherms, I found that there was a mismatch in the scaling of metabolic rate and respiratory surface area. And this is really exemplified when you look at the posterior distributions of the slope values from the model. And so on the top, we have metabolic rate. And so you can see that the slope is quite shallow compared to that of respiratory surface area. But for ectotherms, I found a match in the scaling of metabolic rate and respiratory surface area. And you can see that the posterior distributions here almost are-- the mean effect size is almost identical, and the distributions overlap. And so this suggests that respiratory surface area is not simply a recasting of the differences in metabolic rate between endotherms or ectotherms or is not directly related to the differences in their metabolic rates. So collectively, this work suggests that respiratory surface area plays a critical role in understanding variation in metabolic rate across vertebrates and possibly when we're using the metabolic theory of ecology to predict any life history traits or any other traits or thought patterns and processes that we may want to rope respiratory surface area in. But one problem or caveat with this work is that it's largely agnostic to activity level. And so we know that activity level is correlated with both metabolic rates and respiratory service area. But quantifying activity level on a scale across vertebrates is really difficult because many vertebrates employ different forms of locomotion such as walking or running or flying or swimming. But for fishes, activity level can be quantified through the morphology of the caudal fin. And this is because the caudal fin aspect ratio, which is the squared height of the caudal fin divided by its surface area, is a morphological and quantitative correlate of swimming speed and activity level in fishes. For example, species with high caudal fin aspect ratios like this tuna are faster swimmers and are more active compared to the species with lower caudal fin aspect ratios such as this grouper. And so using such anatomical or morphological measurements is really beneficial when you're working in a comparative or macroecological context because it's often unrealistic for detailed data such as swimming speed or even actual caudal fins from live animals to be available for measurements. And so you can instead turn to field guides and other anatomically correct illustrations such as what's shown here. And this really opens the door to obtaining these measurements for many species, more than what we can gather from fieldwork or direct observations. And so I wanted to add an activity level by measuring caudal fin aspect ratios from these anatomically correct drawings. And so this brings me to the second question where I examine the relationship between shark gill surface area and ecological factors including activity level, habitat, and maximum size. But today, I'm really only going to focus on the relationship with activity level in the interest of time. But you can always check out the paper in Journal of Morphology if you'd like to read more. So one important point that we also encountered with the first question is that like many morphological and physiological traits, gill surface area does not increase at the same rate with body mass throughout an organism's lifetime, so meaning that it's scaled allometrically for most species. And so to understand how gill surface area relates to activity, we have to bring gill surface area into a scaling context and examine how the ontogenetic slope, which biologically would be interpreted as the rate at which gill surface area scales with body mass, and the ontogenetic intercept, which biologically is the gill surface area at a given body mass, relate to activity. And so to dig into this question, I collated gill surface area for as many shark species that had raw data including gill surface area that I directly measured, which we'll talk about a little bit later. And so at the time, there were 12 species that had enough data from multiple individuals to estimate species specific allometries of gill surface area and body mass. I then measured the caudal fin aspect ratio for each of these species using anatomically correct illustrations and then used mixed effects models to estimate the species specific relationship between gill surface area and body mass, which is shown here, and then assesses how these species specific slopes and intercepts vary with caudal fin aspect ratio across species. So for the slope, I found no relationship with activity level such that the slope was the same for each species regardless of activity, meaning that the gill surface area increased at the same rate ontogenetically for species regardless of activity. But the intercept did vary with activity level such that more active species had a higher intercept. And this means that they had a greater gill surface area at a given size for more active species. And so the fact that the slope did not vary with caudal fin aspect ratio but the intercept did vary could be for many reasons, which includes the narrow range of slope values that I found for these 12 species, whereas the intercepts were highly variable and actually ranged over an order of magnitude. And so this isn't actually that surprising because intercepts are known to vary considerably where slopes are more consistent possibly due to geometric constraints, although that's controversial. So it can also be due to the number of sharks that were able to be included in the study, which was based on data availability because we do know that slopes vary a little bit. And so a big part of my PhD was also collecting and measuring gill surface area along with many collaborators and colleagues for as many shark species as I could. And so I have a short video of the sampling and dissection process. And so for most of my fieldwork, I participated in coastal long lining and gill netting off the coast of Florida, Georgia, and South Carolina with other research labs that were doing this research. And I collected shark specimens opportunistically and took them back to the lab for gill surface area dissection and microscopy. And so that's what's happening now. The gill hemibranchs are actually being separated. And so that's what the gill filaments look like. In a shark, there's going to be some on the other side. And then to measure gill surface area, you count the filaments, which is what's happening here. We're bending them so we can keep track. And then we measured the length of the filaments and then extrapolate that to each bin. And then eventually, we're going to cut them out, which is what's happening here. And so this is really delicate, time consuming work. It takes about 30 to 40 hours to measure the gill surface area for one individual. And so after the filaments are cut out, we separate them. It looks like this. And then we actually have to measure the other two components of gill surface area, which are lamellae frequency, which is what is shown here. So those are the lamellae, which is the good stuff. And we measure the number of lamellae per unit distance under a microscope. And so what's happening here. We're cutting the filaments apart, and then we're going to take magnified photos and then use image processing software. So we'd line them up with a ruler because we are lazy and don't calibrate our scopes. And then we take the pictures. And then we measure using image processing software. And then the next step is going to actually be cutting a cross-section of an actual lamellae so we can take another magnified photo such as this and then actually trace that surface area. And so that is the process for measuring gill surface area. And so I used these data to answer the third question where I examine the relationship of gill surface area and life history. And so specifically, I was interested in the relationship among gill surface area growth and maximum size because I wanted to test a central prediction of the gill oxygen limitation theory. So formalized as life history theory, decades of work have revealed that body size and other life history traits that are related to growth, survival, and reproduction are optimized through selection to maximize fitness. And so there's been wide empirical and theoretical support for life history theory more broadly. But the gill oxygen limitation theory turns this idea on its head and instead proposes that in fishes, maximum size and its corresponding growth rate are mechanistically driven by gill surface area such that gill surface area actually limits growth and maximum size in all fishes and other water breathing organisms. And so this goes back to this conceptual figure that you've now seen a few times. And the gill oxygen limitation theory suggests that a reduction in the ratio of gill surface area to body mass, which is on the y-axis, results in a reduction in the relative oxygen supply, which ultimately determines maximum size and thus growth in fishes. And so as I shared with you in the introduction, this theory predicts that climate warming will result in a reduction of the maximum size of fishes. And so this theory has actually been already used in the literature, as well as being taken up by policy to actually predict future changes to fish biomass and distribution. So a central prediction of the gill oxygen limitation theory is that the gill surface area is tightly correlated with growth and maximum size. And specifically, I'm talking about growth in terms of the von Bertalanffy growth model. And so this would be the growth coefficient and asymptotic size estimated by this growth function. So in this model of fish growth, the asymptotic size is the size a fish would reach if it were to grow infinitely. And the growth coefficient is the rate at which this asymptotic size is approached. And so one inherent consideration from this model of fish growth and really life history trade offs more broadly is that growth coefficient and asymptotic sizes are inversely related, such as what you see here. And so this means that a species generally grows either faster to a smaller maximum size or grows more slowly to a larger maximum size. And so if we look at this with actual data where each data point represents a species pair growth coefficient, which is on the y-axis, and asymptotic size on the x-axis, we do see an inverse relationship, although there's quite a bit of variation. And the gill oxygen limitation theory would predict that this variation is solely related to gill surface area. So in other words, the gill surface area can explain this variation in growth performance. And when examining the relationship of gill surface area growth and maximum size, early work on this theory dealt with the inverse relationship between the growth coefficient and asymptotic size by using an index called the growth performance, which is simply these two metrics multiplied by each other and then on a log scale. And this index was used because ideally it would integrate the trade offs that exist between growth and asymptotic size as modeled by the von Bertalanffy growth function like I just showed you. And so this early work also used an index of gill surface area called the gill area index, which is calculated using mean gill surface area and mean body mass data and scaled to how a species gill surface area changes with size. And such an index was used at the time because there's really a lack of raw gill surface area data or such data from multiple individuals of the same species. And so when first examined over 40 years ago across 42 species of fish, only a weak relationship between gill area index and growth performance was identified. But if gills are indeed related to growth performance and maximum size, then this relationship and the gill oxygen limitation theory more broadly has far reaching consequences and implications, particularly in light of increasing temperatures with climate change. But also if there's no relationship, then we should probably rethink how we're predicting changes in life history and ecology based on individual physiology. So I wanted to revisit this relationship because in the 40 years since it was first established, we've seen an increase in the availability of raw gill surface area data and higher quality gill surface area data as well as an incredible advancement in statistical techniques. And so this offers us an opportunity to really dig into the relationship from many angles and understand whether gills are actually related to growth and maximum size in fishes. And so this particular project has several separate research questions that include a deep dive into statistical methods in ecology, re-examining existing theory in data, and a meta analysis on gill surface area growth and maximum size across fishes. I'm only going to focus on a few during this talk in the interest of time. And so it's important to note that to answer these questions, I took two separate but complementary approaches. So I first revisited the original data set that was used to establish this relationship over 40 years ago. And that's this relationship of gill area index and growth performance across 42 species of fish. And I then conducted a meta analysis of gill surface area growth and maximum size to harness the power of more data to examine the generality of this relationship across more species. And so my first question relies on both approaches because here I ask whether examining the relationship between gill area index and growth performance across more species gives us a different result than what was first found, which was really only a weak relationship. And so second, I use the meta analysis data set, which has data for more species, to refine the metrics of gill surface area used. And this allows me to ask whether other ways of looking at gill surface area change your understanding of the relationship between gill surface area and growth. And then finally, I ask whether activity level explains more variation in growth performance compared to gill surface area. So for this first question, looking to see if more data gives us a different answer. Importantly, the original relationship that was first established used a regression method that was really common at the time. But it assumes no directionality between x and y variables. So in this case, no a priori cause and effect between gill area index and growth performance. But we know that the gill oxygen limitation theory does make explicit predictions about the directionality of this relationship because it suggests that gill surface area explains variation in growth performance and that gill surface area actually limits metabolic rate and growth. And so before I compare the relationship of gill area index and growth performance to that with more data, I first needed to flip the axes. And so now growth performance is on the y-axis. It's the response variable. And gill area index is on the x-axis. And it's the predictor variable. And now this relationship is assessing whether gill area index explains variation in growth performance. And so when I estimated this relationship with the axes flipped but using the same data set of those 42 species from the original theory, I found that there was a significant relationship between gill area index and growth performance. So then I collated additional data from data mining from the literature. And I went back to that data set of gill surface area that I and my colleagues measured directly. And this brought the total to 132 species of fish. So I then re-estimated the relationship between gill area index and growth performance across these 132 species. And I found an almost identical relationship to the original relationship where gill area index explained significant variation in growth performance. And so this tells us that examining this relationship with more data did not change our results. But gill area index, while successful at the time this relationship was first examined, does not really capture the known variability of gill surface area and body mass within species. And this is largely because it's estimated with mean data, which we know is not ideal because gill surface area increases with body mass allometrically. So ideally, we would use both the slope and intercept as you saw in the last question to see how these metrics of gill surface area relate to growth performance. And so this is exactly what I did for the second question. And so I took the data set of 132 species. And I took a subset based on those species that had enough raw gill surface area data, which was only 32 species. I then expanded the multilevel modeling framework that I developed for that very first question to be able to include raw data. And so now the first level of this model estimates the species specific ontogenetic allometries of gill surface area and body mass. The next level extracted each species' slope or intercept. And then the final level use these slopes or intercepts as predictors of growth performance across species. And just like when I use this model in the last question, each iteration happens in succession. So it's propagating the uncertainty across all levels of the model. So for the ontogenetic slope, I did not find a strong relationship with growth performance. And this indicated that the slope of the gill surface area allometry or the rate at which gill surface area scales with body size does not explain much variation in growth performance across species. And this was also true for the ontogenetic intercept. So it didn't explain much variation in growth performance across species. So even though neither the slope nor the intercept had strong relationships with growth performance, there was a fairly large amount of variability in these relationships. And so this led me to ask what other variables are related to growth performance? And so this brings me to the last question I'm going to share with you in this last section, which is focused on whether activity level as measured by caudal fin aspect ratio explains more variation in growth performance compared to gill surface area. So I went back to the Bayesian multilevel modeling framework. And I added in caudal fin aspect ratio as a predictor in the last level where I examined whether gill surface area as measured by the slope or the intercept explained variation in growth performance. And so for these results, I'm going to walk through the coefficient plot here. So we're comparing the standardized effect size of the slope or the intercept and caudal fin aspect ratio. And so I found that the caudal fin aspect ratio effect size was significant, or it was not equal to zero, and to explain variation in growth performance. Both the slope of the gill surface area allometry did not. And this was the same for the intercept. So this means that activity level explains more variation in growth performance compared to gill surface area. And so the overall take home from this last question is that the relationship between gill surface area and growth performance across fishes here examined is pretty weak and often nonexistent. And this is especially true when I refined the metrics of gill surface area used in this question. For example, just like the slopes in the intercepts as opposed to the gill area index, which we know doesn't really capture the variability of gill surface area within species. And so if we think about these three questions holistically, I think that this body of work shows an intimate relationship between metabolic rate, respiratory surface area, and activity level both across and within species. But clearly, the relationship with life history traits, specifically growth and maximum size, is much more complicated. But we're working on expanding this framework beyond my PhD work. And we've drilled into the relationship between metabolic rate and life histories across fishes. And we found that only those life history traits that accounted for a trade off such as growth performance were important in explaining variation in metabolic rate rather than traits in isolation such as the growth coefficient or maximum size by itself. And this was a pretty big finding because it's often assumed that metabolic rate correlates pretty closely with life history traits. But this idea has surprisingly rarely been tested across species, particularly for ectotherms and fishes specifically. And so we've also investigated the relationship between gill surface area and other morphological traits such as brain size and gill slit height, which is the height of the gill slits, which would be theoretically and empirically correlated with the cross-sectional area of the head. And we found that individuals that had both larger brains and larger gills especially as they increased in size, which means that a larger gill surface area may fuel the metabolic demand of a larger brain. And then we also found that the scaling of gill surface area and gill slit height was highly correlated across the five species for which we had these data. And a cool result from this paper was that gill slit height that was measured from anatomically correct illustrations and field guides matched that that was measured from physical specimens, which suggests that we may only need to go to books to look for measurements of gill slit height and thus possibly gill surface area if this relationship holds across more species. And this really opens the door to obtaining these measurements for all shark species. We also have some work in progress with some of my colleagues, particularly Tanya Prinzing, who is a Master's student at Simon Fraser University, and Nick Wegner down at NOAA in La Jolla, that's looking at the relationship between metabolic rate and respiratory surface area within the same individuals experimentally. And for this, they've measured gill surface area and metabolic rate in the same individuals for nearly a size range of the horn shark. So that's going to be a really cool study that's going to come out soon. And then lastly, we're working on bringing temperature and oxygen directly into the mix as well as population dynamics. And this will help us to start understanding how physiology may affect fisheries in a changing climate. And so although we're still very far away from the goal of developing simple ways to predict life histories and population dynamics across species and temperatures, this work hopefully provides foundation for doing so. And hopefully, it'll be important considering the speed of environmental change as well as the dwindling funding that's available for conservation and management efforts. But one really challenging aspect of macroecology is identifying the mechanisms that underlie these broad patterns. And it's quite difficult if not nearly impossible to identify the mechanisms underlying these patterns using macroecological approaches alone. And so I wanted to expand my training and gain new perspective by working on more mechanistic-based questions that can be directly tied in with fisheries management. And so while I'm at Alaska Fisheries Science Center, I'm going to be working on predicting Pacific cod spawning habitat suitability in the Eastern Bering Sea. And so specifically, I'll be coupling model bottom temperatures from the Bering Sea 10K ROMS model with an experimentally derived relationship between hatch success and temperature to examine how the availability of thermally suitable spawning habitat for Pacific cod have shifted across space and time in the Eastern Bering Sea. And before I end, I just wanted to take a moment to thank all who made this work possible, which you can see is a very, very long list. And I will take questions if we have some time. HEATHER TABISOLA: Awesome. Thank you, Jennifer. [CLAPPING] The clap for everybody that doesn't get to since we're not in person. All right. I just invite folks to go ahead and type your questions in the chat. Or you could feel free to also jump in on video. Just make sure you put your name in the chat so we have a queue that we can go down. I'm looking forward to hearing how things go with your project. JENNIFER BIGMAN: Thank you. HEATHER TABISOLA: Yeah. No questions from anybody? Oh, my goodness. People must be tired. OK, so tell me-- oh, here we go. Thank you, Ben. All right. Does total gill mass equal surface area? JENNIFER BIGMAN: That's a great question. So that's actually not been looked at too often. It has been looked at, but really in a few species I believe, maybe one or two. And it actually scales one to one with body mass as you would expect because it's gill mass. So it doesn't actually scale at the same rate as gill surface area. But this is really more anecdotal that I've only seen in one or two species. And that's probably because it's difficult to measure gill mass because at least for sharks, you'd have to remove all of the filaments from the septum and be a little easier to measure in [INAUDIBLE]. But that hasn't really been done. But for more than one or two species. HEATHER TABISOLA: Ben says, thanks. I was impressed that you said 30 to 40 hours for each specimen. JENNIFER BIGMAN: I have lots of help. [LAUGHS] HEATHER TABISOLA: Still a lot. [LAUGHS] Any other questions from folks? OK. No questions. All right. So Eva Schemmel says, any thoughts on assessing energy diverted to reproduction versus growth and teasing out that relationship? JENNIFER BIGMAN: So that's a great question. And that's really a hot topic when it comes to thinking about gill surface area and the gill oxygen limitation theory because there's a current kind of debate in the literature right now that is what causes fish to mature. And the gill oxygen limitation theory-- I think Daniel Pauly, who kind of developed this theory, he recently published a paper maybe just a month ago on why fish mature when they do. And so he argues that fish become oxygen limited, and that's why they mature when they do because of the gill surface area constraint. Obviously, that's not aligned with life history theory. But to look at that, you could throw reproduction and possibly total reproductive output into the last project of mine, which I think someone else that we're working with is going to look at. But it'd be really interesting to see if reproduction is related to gill surface area. HEATHER TABISOLA: Thomas Hurst says, have you or others looked at these relationships between populations? Would that provide a different perspective than cross species? JENNIFER BIGMAN: Right. So that's also a great comment. We have not mostly because these data are pretty rare. And so it's hard to get data from multiple species, not least multiple populations within species. Now, we have looked at gill surface area in black tip sharks between the Atlantic and the Gulf of Mexico. And we didn't really find a difference, although we didn't have a great overlapping size range just because logistically, getting small and big animals in both ocean basins was difficult. But perhaps this could be done in [INAUDIBLE] if there's more data. HEATHER TABISOLA: And Matt Wilson says, nice presentation. Thoughts unexpected biological/ecological responses of fishes to ocean warming related to deoxygenation? JENNIFER BIGMAN: Right. So I guess that's the whole idea behind the gill oxygen limitation theory, or that's what it's being applied for to use is that basically-- I guess if you think about the balance of oxygen supply and demand with warming temperatures that there won't be enough oxygen to meet the demand. And so fish will shrink. And that's kind of the whole idea of gill oxygen limitation theory. And so we don't really know if that is mechanistically supported. And that's kind of what we're all trying to test. But there is a decrease in fish size and shifting distributions with ocean warming. Whether that's related to oxygen and gills specifically, we're not exactly sure yet. HEATHER TABISOLA: I will give folks one last chance for questions. And if not, I'll take the opportunity just to remind folks that we will be here next week with Lauren, with Dr. Lauren Rogers. Lauren says, great talk. Everybody's saying great talk. So there you go. And if you have further questions or a slide, please reach out to Jennifer directly. Dave has a question. Dave Kimmel. Areas of hypoxic water have increased over the past decade. Is there an impact on gill surface area of fish in these areas that you are aware of? JENNIFER BIGMAN: So again, that's another good question. So there's a few different I guess schools of thought. What we actually see in hypoxic areas or in the lab when oxygen is lowered is that gills actually have quite a bit of plasticity. And so, for example, in carp there's these things called interlamellar masses that slough off and expose more lamellae. And so I guess the argument is that gills provide enough surface area to diffuse oxygen over the surface even in hypoxic waters because they can either slough off these interlamellar masses, which we don't know if they exist in all species. And then secondly, they can increase the vascularization of their gills. So generally, when fish are swimming around doing their daily thing, the entire gill surface area is not vascularized. And so you can increase that vascularization. And so there's a controversy right now in the literature whether the hypoxia or lower oxygen is going to be enough to push those fish over the edge. Or do fish already have enough adaptations to be able to handle that added stress? And so I guess the answer is we don't really know. HEATHER TABISOLA: Thanks from David. All right, you guys. I think-- oh, Deana. Deana. Can you define or elaborate on residual respiratory surface area for me? JENNIFER BIGMAN: Right. So when I was working on this first chapter of my thesis, we were trying to come up with a way where we could look to see if respiratory surface area explain variation in metabolic rate. But because this is macroecology, and we're using mean trade values, and the metabolic rate and respiratory surface area is measured on different sets of species opportunistically-- so I'm going through the literature and finding values-- the body masses don't overlap. And so you can't just plug that body mass into your model. And so I had to come up with a way to account for both body masses. And so that's how we came up with residual respiratory surface area. And this is actually a common technique in macroecology to use residuals when you have to trace the scale of body size. However, residual respiratory surface area is just one measure of the respiratory surface area after the body size has been accounted for. And so hopefully, I was able to combat a little bit of that problem by plugging it into a Bayesian model and actually propagating the uncertainty across all levels so I'm not just dealing with one residual respiratory surface area estimate for each species, but it's a distribution of possible values. And so I guess more simply, residual respiratory surface area is just the respiratory surface area after the measurement body mass has been accounted for. And then we plug that into the model. HEATHER TABISOLA: Awesome. Thanks [INAUDIBLE]. Does that answer it, Deana? Cool. OK. All right. I think with that, we have five minutes to 11 o'clock. And we will close out seminar for this week. And Lauren reminded me next week is Thanksgiving. So we have break week. [LAUGHS] I know. It's not-- yeah. We will resume the week after with seminar. And then Phyllis Stabeno will close out the season. So thank you everybody for joining us. Thank you, Jennifer, for being here. And hopefully, one day we will actually get to meet in person. So thank you, everyone. JENNIFER BIGMAN: Thank you for having me.