Winifred Frick (University of California, Santa Cruz): White Nose Syndrome
Hi. My name is Winifred Frick, and I’m an Assistant Adjunct Professor in the Department of Ecology and Evolutionary Biology at the University of California in Santa Cruz. I’m gonna be talking today about my research that I did a couple years ago on the impacts of a disease called white-nose syndrome on a bat, the little brown bat, that used to be really common in the northeastern United States. White-nose syndrome is a new disease that was only discovered a couple of years ago. And it’s caused by a fungus that grows on bats while they’re hibernating. It was first discovered, actually in 2007, when researchers from the New York Department of Environmental Conservation went in to do their annual surveys of hibernating bats. So, biologists go in, and the bats in the northeastern US aggregate during the winter into what we call hibernacula. And these are caves and mines where bats spend the winter. And biologists from different states go in and count them every year to see how many are there and to monitor them. And part of the motivation for this is because we have a species called the Indiana bat that is federally endangered… on the endangered species list, so people need to go in and count so we can monitor how those species are being… how they’re faring. And in 2007, people noticed something they had never seen before, which is a bunch of bats dead on the cave floor, and this funny white fungus growing on their faces. It was later that cavers actually came forward with a picture of bats with white fungus on their… white fluffy stuff on their faces, in 2006. So, we can date the introduction of white-nose syndrome to 2006, but we didn’t know it was a problem until 2007. The main concern with white-nose syndrome is that it’s causing massive mortality of hibernating bats in their hibernacula during the wintertime. So, the bats go into these places, and they spend all winter there, in hibernation. They actually turn their metabolisms off and spend the whole winter, and they’re what we call torpid. So, they’re mammals, like you and me, but they actually will turn their… turn their… turn their bodies off to be able to survive the whole winter. And during that time, this is when this… what we now know is a fungus grows on them. And from what we can understand — we’re still trying to figure out all the mechanisms of why this disease is killing all the bats — it’s causing them to wake up too frequently while they’re hibernating, and that makes them use up their fat reserves, and then they die before spring comes. So, this picture here, these different pictures here that I’m showing you, are piles, literally piles, of dead bat carcasses on cave floors. This is from a cave in Vermont. So, the symptoms of white-nose syndrome are this characteristic fungal growth on the face of the bats. This is how it got its name, white-nose syndrome, but actually the fungus grows all over their wings and tail membranes. Also, the bats start doing weird things in the winter. They should be hunkered down and hibernating all winter long, but instead they’re getting out, flying about… flying out on the landscape. Back in 2007, there were all these reports of people saying that they were seeing bats flying out into their backyards in the middle… even when there was snow on the ground, and this is something that these little critters should not be doing. And then in springtime, if the bats manage to survive, we see that they have these terrible wing lesions. And basically the fungus has been sort of eating at their wings all winter long, and they have all of these terrible injuries. Well, it took a couple years to figure out exactly what was causing white-nose syndrome. It was… it was originally named a syndrome because we didn’t know what was going on. We didn’t know what that white fluffy stuff was on their faces. We now know that that is caused by a pathogenic fungus called Geomyces destructans. The fungus wasn’t known to science previously, so it’s newly described. And so David Blehert at the National Wildlife Health Center and others aptly named it Geomyces destructans because of the destruction that it’s wreaking on our… on our bat faunas. And so what the fungus does is it actually grows into the skin issues. It invades the different skin layers. It’s sort of like a really terrible case of athlete’s foot. But the bats really depend on their… on their skin tissues for all sorts of physiological processes, and… and as I said previously, we’re still working out some of the mechanisms of death. It may be that this fungus just irritates them, and that’s what wakes them up. Or it could be that the fungus causes them to become dehydrated, and they’ve got to wake up to try to find water. Scientists are still working out some of those questions. But we do know that they wake up too frequently, and that they starve to death before… before spring comes. So, the motivation for the study that I’m talking about today… and you have to remember that this was… we started this study back in 2008 and 2009, when we didn’t really know very much about white-nose syndrome at all. We just knew that lots of bats were dying. And so what we wanted to know was, were populations in the northeastern United States… because that’s where white-nose syndrome hit… it started… it was first discovered in Albany, New York, in upstate New York… were these populations okay before WNS hit? Were they… were… there were some hypotheses about… well, perhaps the bats were being stressed by pesticide use or depleted insect populations, and that was causing them to go into their hibernacula in more stressed conditions, and that perhaps was causing some sort of secondary infection to emerge. So, we wanted to know, how were these populations doing before white-nose syndrome hit? And then we wanted to know, what was going to be the impact of all this mortality on this regional population of this bat that was really common? Little brown… little brown bats are the bat that’s in people’s attics or in their barns. They’re really quite a common species up in the northeastern United States. And so we wanted to ask this question of, what is the potential impact of this mortality to this population? Do we need to be worried about the persistence of this species? So, to do that, the study was sort of conducted in kind of two different stages, really. The first involved, actually, the use of a data set that my co-author, Scott Reynolds, collected for 17 years. You can see here this barn. This is the Peterborough barn in Peterborough, New Hampshire. And it has a colony… had a colony of little brown bats roosting in the attic. And so, as part of his PhD project, when he was a PhD student with Tom Kunz at Boston University, he’d go to this barn and he’d capture the bats. Here, I have a picture of a mist net. And the bat gets captured in a mist net. You put the mist nets around, and bats get tangled in them. It doesn’t hurt them, but they get tangled, and then you can take the bat out. And then you can put a little forearm band — it’s just a little a little piece of metal that you can squeeze around the forearm; you can see it here — onto the bat. And it has a unique ID. And then you let the bat go, and it lives out its life in the barn. And you go back every year, and you try to capture the bats each year. And you find out… you basically develop a dataset that has this record of how long the bats live for. And we were able to use that over this 17-year time frame to estimate what the probability that a bat lives from year to year is. And we call that annual survival rates. And we were also able to estimate the fecundity, or the number of bats… of young that adult females produce. And as well as the probability that the female bats come back to this barn to breed. So, this gave us quite a bit of information about the… what we call the population dynamics of the bats. So, we were able to use that information. In addition, as I mentioned, state biologists go in and they count the bats in their winter hibernacula, in these caves and mines where bats spend the winter. So, we have a record of how many bats are at these sites from year to year. And we can also use that information to try to estimate what the population dynamics are. How stable are the populations? Are there fewer bats every year at these caves and mines? Are there more bats every year at these caves and mines? Or are there roughly the same number from year to year? And so, we kind of combine these two pieces of information, because the demographic data — the data from that that bat banding study — gives us a really in-depth look, but we only had it from one barn. So, we didn’t know if that would be representative of the population… the regional population as a whole. Whereas these data from the winter hibernacula — these count data — are a little bit more of a rough estimate of how the population is doing. It’s not as precise of information, as detailed of information, yet we had it for 22 different hibernacula across the northeastern US. So, we had kind of a broader regional picture. So, combining those two datasets… so, looking at the demographic rates from the 17 year… what we call a mark-recapture because we mark the bats and then we recapture the bats… from this maternity colony of little brown bats, we estimate annual survival for adults and juveniles, as well as the breeding probabilities. And then from the hibernacula counts, we were able to estimate population growth rates. So… over the past 30 years… because they’ve been doing these surveys for a long time. So, first I’ll just show you a little bit of the data. This is actually data that didn’t appear in the Science paper, but is in a previous paper, but we used it in the Science paper. What this shows us is… on the x-axis, we have the year. So, this is just the range of years that the study took place over. So, Scott started banding the bats back in 1993, and then we… he did the study up through 2008. And then on the y-axis, we have the annual survival probability of adult female bats. So, this is the probability that an adult female bat lives from year to year. And what this graph shows us is a couple things. One is that, overall, there’s a… there’s a fair bit of variability in the survival of female bats. It ranges from about 0.6 — so, a 60% chance of living from one year to the next — to about 0.9. And the other thing is that we found in this study that the chance that you survive from year to year is actually strongly correlated with the amount of rainfall that you get in the summertime in New Hampshire near this… near this site. So, that’s what this dashed line is, is the amount of cumulative summer precipitation, the amount of rainfall that happens in the summer. And so, what we found was that in years that had high amounts of summer rainfall that we have the highest survival rates. And in this study, we hypothesized that that was probably true… because this pattern actually has been found for passerine birds, little songbirds that also eat insects… is that in the northeastern US when you have a lot of rainfall in the summer you have more insects, and so there’s more food, and so the bats probably do better in those years. So, we used the data, here, from… with all this information about how… not just what the actual estimates of survival are, but the variability around those estimates, how much difference there is from year to year in these estimates, to put into our population models. This part here is kind of a… a little sort of diagram, if you will, of thinking about the life of a bat. So, you start off as a… as a juvenile. So, we use the term juvenile to mean a bat that’s born in a… in a given year, and that whole first year of life. So, from the time that you’re born… and bats are born in like June/July… they’re born in the summer… that whole first year until the next June or July. And so your probability of becoming an adult is represented by this S-sub-j. That means the survival of being a juvenile. So, that’s the chance that you make it from being… from the time you were born to when you come back to the breeding colony as an adult the next year. And then there’s the probability, if you’re an adult, that you’ll survive the next year and… or, you know, survive each year as an adult to become… you stay an adult, but it’s your chance that you survive from year to year. Then, if you’re an adult, there’s the chance that you will return to the breeding colony and breed. And so that’s that B-of-a times F. So, that’s your… the B-of-a is the chance that you return to the breeding colony, and then F is, if you return, how likely you are to actually have a pup. And so… and since you can breed after your first year of life, the juveniles also get that probability as well, but it can be different for them from the adults. So, this is just sort of a cartoon about the way we can sort of break down the different stages in life for a bat. And then we can try to do some math to sort of estimate what these different probabilities are, and put those together in a matrix, and do some matrix algebra to determine what the population growth rate will be for these populations. So, that’s what we did. And from the 17-year mark-recapture dataset, we found that overall… with those with those survival rates and breeding probabilities that we had estimated from banding the bats in the barn… that those populations were doin’ okay. They were… what we call lambda, which is just the mathematical term for population growth… it was positive, meaning those populations were actually producing more individuals over time than they were losing. Okay. So, in lambda… if lambda is 1, then… if lambda is 1, then population… it basically means it’s stably the same number of individuals in the next year as you had the first year. So, any number greater than 1 means the population is increasing over time. And any number less than 1 means the population is decreasing over time. And then we wanted… like I mentioned, we wanted to relate how that detailed demographic data related to the regional picture of the populations. And so we did that using these annual winter counts, and we estimated with the annual counts what that lambda with that population growth rate was. And for 86% of those populations, the [population] was stable or positive. And so that kind of told us that… and the regional average was 1.07. And so those numbers aren’t exactly the same, but what it told us was that those demographic rates that we got from the banding study were roughly representative of what was going on across the region. And that gave us confidence that we could go ahead and use those in our stochastic population model, that I’ll talk about it in a second, to try to estimate what the impacts of white-nose syndrome will be on the population. And importantly, the… one of the big take-homes from this was that, prior to white-nose syndrome, these populations in the northeastern US were healthy and growing. These were… and that makes sense. This is a common bat. It’s a bat that can live in your barn and your attic. And so it was… these populations were not suffering. So, by the time we had gotten all these numbers and crunched all this down, we sort of already knew that what was killing the bats was the fungus. But this sort of corroborated that the fungus wasn’t just opportunistically taking bat’s lives because they were stressed from other sources. These populations were actually doing well. And this was… the fungus kind of came in as a novel pathogen, as something from somewhere else. And we now know it came from Europe. At the time, we didn’t know that. So, as in all science, you want to sort of stay within the scope of your data. And so this is kind of to remind us that, although at this point… so, this map is outdated. This map was in the paper, but white-nose has been continuing to spread and continuing to sort of march its devastation across the US and further into Canada. But at the time that we were working on this, the data that we had was from where white-nose syndrome originated, which was in… sort of centered in New York, and sort of in the northeastern US. So, the… when we… when I get to talking about the impacts of regional extinction, it’s important to remember that is regional, and it’s based off of the data that we had, which comes from a particular area. Because little brown bats, as this map down here shows, actually range all the way to the western US and into Alaska. So, here we have graphs that show us actually what we call the raw data, which is actually the… just the count, the number of bats that are counted at a hibernacula in the winter. That’s on our… on our y-axis. And again, year is on our x-axis. So, this is just sort of through time. And so our timeline actually spans from the early to mid-1980s all the way up until 2010. It’s broken into three different panels here because hibernacula come in all different sizes, and so it’s hard to actually see what’s happening at the small… at the small hibernacula if you plot them on the same graph as the large hibernacula. So, this large panel… this panel here that says large are hibernacula they range from 30,000 bats down to 5,000 bats. And here we have the medium. So, this is the hibernacula that range from 4,000 bats down to 1,000 bats. And here we have small hibernacula that range from 1,200 bats down to just a couple hundred bats. And these are just the number of bats counted each year. And so a couple of things that you can kind of tell from these graphs is that prior to white-nose syndrome… white-nose syndrome is… when it enters the system is noted in these… in these gray bars. The bats… overall these populations were increasing, which you sort of can just visually see, and I showed you the math that sort of shows that that indeed is the case. Occasionally, you have a big catastrophic decline, like here, where you’ll see a big… a year with a big decline in bats. And those were usually… those were always well documented in the field notes of the biologists who would go out. And they’d say things like, there was a big flood, or, you know, part of the cave caved in. And so you occasionally do get catastrophic events, but the overall picture is one of populations that were actually doing quite well. But then, in the gray bars, what you see… and you can use this middle panel as an example… that the populations just crashed down when… once white-nose syndrome hits. So, at the time that we were doing this, there were only three years post white-nose syndrome infection. And we did have just a smidgen of data that indicated that perhaps the impacts of this might lessen over time. And there’s quite a bit of work done in disease ecology theory that talks about differences between the way that pathogens or disease are transmitted to individuals, and how that impacts the long-term dynamics of a disease. And we didn’t have the information we needed to do sort of a proper disease model to understand that. So, we used the available data that we had. And what we had was that in… the declines in the first year post infection were very high, averaging around 80%. And then in the second year, there was a little bit more variability, and overall a little bit lower. So, averaging around 60% or so. And then in the third year, we only had two data points, but it was even lower, and it averaged around 45%. So, we took that data… this is actually the same data, it’s just the axis is flipped now, so that we express it in terms of population growth, that lambda. And so 1, again, would be where populations are stable, the same number of individuals from year to year. And so, what we did was we said, okay, well we don’t really know what’s gonna happen in the future. We’re not fortune tellers here. But we’re just gonna take some guesses and build sort of a best case/worst case scenario. So, our best case scenario was that the populations would continue to stabilize. And we could say, okay, we think they’ll stabilize around 1, and we just sort of fit a line that said, okay, let’s go to 1, and let’s see how long it takes us to get there. And it takes about 16 years to get there. And then… so that became our sort of best case scenario. And then our worst case scenario was that they just would stay at that 45% decline rate forever. So, those gave us our sort of best case and worst case scenarios. So, then we do some more math. And this is that same matrix, population matrix, that I showed you on the earlier slide, where we had the cartoon of the different stages that a bat goes through. And this is just expressed in the matrix… a matrix form. So, you start off in one time step with the number of individuals that are a juvenile and the number of individuals that are an adult. And then you multiply that by this matrix, which are those different transition probabilities, that transition of being… a juvenile bat becoming an adult bat, or an adult bat breeding, and the juvenile bat breeding. And that then gets you to… being the number of juveniles and the number of adults in the next time step. So, we do that for susceptible bats, which are bats who haven’t gotten disease yet. And then that feeds into the number of infected bats, based off of the rate of spread, which we calculated directly from the number of hibernacula that were becoming infected each new year. And so, the same… so, the… these transition values for the susceptibles came directly from the data from that banding study, the mark-recapture study. But then, when the bats become infected, we have to change their survival probabilities, right?, because their chance of surviving from white-nose syndrome is much lower. And so we based those off of the observed population declines, and we used these different scenarios to sort of say, well, what if decline sustained at 45%, or what if declines started to ameliorate, or the population started to stabilize over time, but then persisted at a certain… with a certain level of impact from the disease. So, the point of this exercise is not necessarily to sort of fortune-tell exactly what was going to happen, but to look at a range of scenarios so we could understand the scope of the impact that the observed mortality that we were having would have on the populations. So… and then this fancy term stochastic, that comes into play in that we let the… those… all those probabilities vary. And so we basically, you know, put the data in, and we… and we run it out in a bunch of different simulations, forward through time, and we let things vary a little bit, and that adds a certain amount of noise. So, stochastic just means random. So that we could… allow sort of for chance events to happen as well. So, we had to have some assumptions in the model. So, we had to pick a starting population size. We don’t really know how many bats there are in the northeastern US, so we used the records of how many bats had been accounted for in hibernacula. But we know that there have to be more than that, but we don’t know how many more there have to be. We don’t know what percentage of the hibernacula that we know about is of the total population. So, we had to just kind of take a guess. So, because of that, we actually fit… we actually stated that what we would call extinction in the model is when the number of bats remaining in the population was only 0.01% of the starting population. And that would mean that there would only be 650 bats left across that whole five-state northeastern US region that I showed you in the map. And so our idea here was that would be basically functional extinction, that there weren’t enough bats left on the landscape to be considered a viable population. We also assumed that the population was single and well-mixed, so there wasn’t a lot of spatial structure, so that… basically, bats have every chance of being in different places across that range. So then, we did a thousand simulations, out 100 years into the future with these different parameters, to see… to calculate what the probability of extinction would be based off of these numbers. And that’s what these graphs show us. So, the first scenario is this… what if declines continue at this 45% rate, meaning that the… for your first year, once you get infected, you get… you decline by around 80%, and the second year you decline by 65%, and then the third year declines by 45%, and every year after you continue to decline at that 45% rate. And that the spread continues at the rate that we’ve been seeing it. And so what you get is this horrifying prediction that within about 16 years all of the bats go extinct. So, we reach that extinction threshold incredibly rapidly. And if we say, okay, well, we don’t… maybe there’s some chance that the populations are gonna stabilize. And so if they stabilize at a 20% decline — that the disease only ends up causing a persistent 20% decline each year — then this still goes to extinction, but it takes a little longer. It’s now, you know, between 20 and 30 years. And likewise, if you drop down to only a 10% persistent effective mortality from the disease, you still… eventually the populations go extinct, but it takes us closer to 60 years. And then it only a 5% persistent mortality, you get to about 60% of the bats… there’s a six… sorry, there’s a 60% chance that the bats will go extinct over 100 years. Now, in these what we call… this kind of exercise is called a population viability analysis, and 100 years is typically the time frame that we use for assessing the risk of extinction. And under normal circumstances, a prediction that there’s a 60% chance that a population may go extinct in 100 years would be considered a dire prediction, and that this would be a species that vastly warranted being covered under the Endangered Species Act. It only looks good here because of the sort of horrifying first… implications of the… of the worst case scenarios. And it isn’t until you get to only a 2% persistent effective mortality that you see an appreciable drop in the probability of regional extinction. And under all scenarios, we see a significant population crash, rapidly, even at that 2% level. So… and that’s… and that’s been played out. That is what we have seen. There has been a massive… a massive collapse. And so there’s other ecological ramifications for losing… even if you don’t lose the species, losing the abundance of that species, and what the kind of impacts that may play out in the ecosystem are. So, the conclusions from this particular work were that white-nose syndrome is threatening imminent regional extinction of the little brown myotis, the little brown bat, that was once formerly quite common in the northeast. And that even if populations stabilize, or the declines ameliorate over time, sustained mortality from white-nose syndrome has to get to less than 5% per year to significantly reduce that probability of extinction over a 65-year time frame. So… but the other take-home is that before this disease was introduced, before the fungus was introduced into North America, these populations were doing really well. They were robust and growing. They had been hard hit in the 70s, likely from DDT. But they were… they were experiencing strong population growth. So, you know, if we can find ways to control the mortality and the spread, these populations do stand a good chance of coming back and recovering from this. Now, this work was published a couple years ago, and we’ve learned a lot, and we’re continuing to learn a lot about white-nose syndrome. So, it’s… I encourage you to go and look at current research, or access things online to see what is the current status of what’s happening. We have a big research program going right now to look at the spread of white-nose syndrome across the continental United States, and to look at the seasonality of the disease dynamics. We’re really sort of diving into trying to understand the ecology of the transmission dynamics in the spread of white-nose syndrome, and what kind of impacts that has on populations. And we’re focusing on the other species that white-nose syndrome affects. So, all the work that I’ve been talking about today focused on the little brown myotis, or the little brown bat, but there are a number of other species — about six other hibernating bat species — that are affected by white-nose syndrome right now. So, we’re looking into transmission dynamics, and spread, and other… and other types of impacts. Well, thank you so much for listening. And in… to close, I’d like to acknowledge my co-authors and collaborators who worked very hard on this project with me, and also to the United States Fish and Wildlife Service, who funded this work. And I’d also like to give a special thanks to the countless individuals and the state biologists who have done an extraordinary amount of work over the past several decades, really, of going into these icy-cold hibernacula and counting bats year in and year out, and that gave us the ability to do this analysis. And without that, and their detailed record-keeping and their detailed notes, none of this would have been possible. And they’re really at the forefront of fighting for the bats, and for fighting for saving the bats from this terrible disease.