Aquatic Ecosystem Vulnerability to Fire and Climate Change

Welcome from the US Fish & Wildlife Service’s National Conservation Training Center in Shepherdstown, West Virginia. My name is John Ossanna, and I would like to welcome you to our webinar series held in partnership with the US Geological Survey’s National Climate Adaptation Science Center. Today’s webinar is titled, “Aquatic Ecosystem Vulnerability to Fire and Climate Change in Alaskan Boreal Forests.” We’re excited to have Stephen Gray and Jeff Falke from the US Geological Survey with us today. To introduce our speakers, we have Emily Fort, who is currently serving as the acting USGS Director of the Southeast Climate Adaptation Science Center. Emily? Thanks, John. Hello, everybody, and welcome. Thanks so much for your time. I’m happy to introduce our speakers. We have Jeff Falke. He’s a research fisheries biologist with the US Geological Survey and the acting unit leader of the Alaska Cooperative Fish and Wildlife Research Unit at the University of Alaska Fairbanks. Jeff holds faculty appointments as assistant professor in the Department of Fisheries, Institute of Arctic Biology, and the Department of Fisheries, all at UAF. He’s worked on issues surrounding the interaction between fire, climate change, and aquatic ecosystems since 2010. Steve Gray is the USGS director for the Department of the Interior’s Alaska Climate Adaptation Science Center. He has a great deal of experience in this arena, as well. Thanks, Steve and Jeff, for your presentation. Thank you, Emily. Stephen, you now have the floor to give your presentation. Thanks so much. Thank you, John and Emily. Let me know if you see my screen coming up, and we’ll go from there. You’re all good. Thanks, everyone, and thanks also to Scott Rupp, who could not be here today, our university director from the University of Alaska Fairbanks, got called away at the last minute to DC to attend to some business, there. I’m really just stepping in for him. What I wanted to do first is take a few minutes to set the stage and explain how this work that you’re going to be hearing about today fits into the larger Alaska Climate Adaptation Science Center portfolio. Jeff’s work is not a CASC-funded project but, rather, it leverages past and ongoing work at the CASC. In collaboration with our university leadership, we’re looking more and more to this model where the CASC provides basic building blocks of a project that then brings in significant support from other sources. We’re already seeing this happen in some projects that are funded by NASA, a couple of large NSF EPSCoR efforts and more and more work with DoD, Department of Defense. It’s a win for everyone, because the initial CASC investment attracts new funding for the university and our USGS partners. We all get the additional benefit of the capacity that gives us for addressing key natural resource management issues. Again, just wanted to set the stage there. One of the first efforts that you’ll see leveraged in here is the CSC ongoing work related to climate projection downscaling. We’ve been at this for over seven years. Unlike many similar efforts in in other regions, we’re not primarily focused on just increasing the spatial resolution of climate projections, but rather we’re focused instead on providing management-relevant climate information. That often means filling in gaps in the existing monitoring network or the historical climate record that we have. Much of that work relates to creating datasets related to snowfall, because that’s so important to the work that’s done here, and also plays into some of the fire work that’s been done. Also, we’re really doing this in a way where we’re bringing information or input from the user community to guide this effort. Don’t worry about the exact contents here, but you have a list of target variables that have been developed in collaboration or cooperation between the climate modelers, the ecosystem modelers, and the resource managers. These are basically climate variables that they need to understand the systems that they’re working on. That’s what’s guiding this effort. Second part of the previous work that plays a role in what Jeff will be presenting today is the work of our integrated ecosystem modeling effort. This is work that brings together three existing models to describe portions of processes of this Boreal in far northern ecosystems in a way that each bit or component of the model can talk to the others. It makes sense. It’s a real-world depiction of the way that the systems operate. A lot of what you’re going to be hearing about today builds on or relates to output from that model that speak specifically to characteristics of burns or fires in these northern ecosystems. More than that, this is just an overall piece of the vision of how this program and the operation of the CASC, more broadly, fits together. I already told you about how we’re developing these climate products that are serving as inputs to the modeling effort. What Jeff is going to be talking about today is this key component of taking those model outputs and data sets, and treating that in a way that helps us gain value in terms of applying it to management decisions. I think I’ll leave it right there and just pass it over to Jeff. Thanks for the opportunity to be here today. Thanks a lot. Thanks, Steve, for that introduction. That was great. To start out today, I’d like to thank my co- investigators on this project, Scott Rupp, Peter Bieniek and Helene Genet. Also, the project personnel that are working with me on this project, postdoc Stephen Klobucar, project biologist Deanna Klobucar, and a PhD student, Elizabeth Hinkle. This work wouldn’t have been possible without the funding and a network of collaborators that we’ve put together as part of this project. The project is primarily funded by the Department of Defense and their strategic environmental research and development programs, specifically their resource conservation and resiliency program area. We’ve also built a network of collaborators to work with on this project, including USGS and UAF, as well as Fish and Wildlife Service, the Alaska Department of Fish and Game, and then the Alaska Fire Science Consortium, and the Northwest Boreal LCC. Some background on the project. We know by now that across the western US that wildfires are increasing in frequency and magnitude. This is also linked to climate. We know that in large fire years, they’re also warm and dry. There’s a direct link to climate there. Also, large fire years typically occur in years that have an earlier snow melt. In combination with the shorter winters, the fire season has increased dramatically over the past several decades. What about the future? From a global perspective, fire dynamics are expected to continue to change over both the near term and the long term. We expect changes in the increases in the likelihood of fire into the future ramping up with changes in climate. If you look…It might be a little bit difficult to see here, but across the northern latitudes, there’s strong agreement in the direction of this change based on a suite of climate model. Now, fire is widespread across interior Alaska. Here, the red polygons are the perimeters of fires that have occurred in interior Alaska since the 1940s. This basically matches up with a distribution of the boreal ecosystem in interior Alaska. This is because fire is really the primary disturbance agent in this ecosystem. Similar to the western US, fire dynamics are also changing in Alaska. Here’s a graph of the total area burned across the state from 1950 to present. What we see here is that fires that have been classified, or fire years that have been classified as much above normal or above normal have been increasing both in magnitude and frequency since about the 1980s. Some of this is likely due to the changing environment that’s occurring in Alaska. Here’s some data from there recently published National Climate Assessment. Basically, Alaska is warming at about twice the rate of the lower 48 and the rest of the US. Looking out into the future, we expect a warmer environment in Alaska, regardless of the emission scenario that you look at. Predictions range some increases in mean annual temperatures anywhere from 8 degrees Fahrenheit all the way up to 16 degrees Fahrenheit, especially pronounced in the Arctic, in this sub-Arctic regions of Alaska. What does this mean for ecosystems? Well, climate change is predicted to have major effects on terrestrial ecosystems, mainly through vegetation change. For example, here on the left is a graph showing the predicted distribution of spruce and deciduous forests in Alaska under two different climate change scenarios. We see spruce forests decreasing, and deciduous forests increasing over time. The other major player here, of course, is permafrost. Here’s a map the current distribution of permafrost. The blue areas are places where permafrost is present, and black is absent. Once we get out to about 2100, the predictions are showing that permafrost will be largely absent from interior Alaska. Of course, fire can exacerbate vegetation change and permafrost stock, and then vice versa. There’s feedback between these processes. This project that I’m going to do tell you about today, we’re just getting started on. It deals with aquatic ecosystems response to fire in climate change. I thought I’d give a little review on the fire effects on aquatic species and habitats just to get started here. The question I get asked a lot is are fires bad for fish. Well, according to Smokey the Bear, apparently so. There’s an old poster. I don’t know if you can read this header on the poster, but it says forest fires catch fish, too. That can be the case. High magnitude disturbances such as fire can have immediate impact, including species extirpations, especially when they’re exacerbated by other events such as debrief flows and floods. It’s increasingly understood that for fish and aquatic ecosystem, that fires aren’t necessarily a bad thing for aquatics. In fact, periodic fires can contribute to the long- term persistence of fish populations through these disturbance processes that create and maintain critical habitats. This is particularly important in smaller streams, because the instream processes are really closely linked to the surrounding landscape. Fire can affect various aspects of stream ecosystems differentially. I’m not going to go through each of these. Fire can affect channel stability, can increase discharge and flooding, increase the input of large wood in the streams, increase settlement and turbidity, and then, of course, increase solar radiation getting to the stream if the riparian area is burned. The rate of ecosystem recovery from fire also depends on the metric that you’re looking at. Here’s an older figure from a paper by Bob Gresswell, where we have a bunch of different stream ecosystem components and their response post-fire, in the year post-fire, and then off into the future. The main point here is that the rate of recovery of these processes varies immediately following the fire versus kind of further out into the future. We can see here, for example, that they’re spikes and turbidity and nutrient falling fire which quickly tail off. It’s expected that the organisms such as fishes recover more slowly, but we’ve seen that in many cases that these highly mobile organisms like fishes are often the first species to recolonize into these burned environments following a disturbance. With respect to the biological effects, there’s not a huge body of literature out there on the subject. If we look across all of it, the responses of fish in other aquatic attacks and fire seem to vary by trophic level, and as a function of the intensity of the fire, and then how long it’s been since the fire. Pulses of nutrients following a fire can increase primary production in streams particularly if the riparian canopy is removed by fire, but this is mediated by increases in turbidity. The abundance and quality of fish and birds of prey can also be affected, especially when this riparian vegetation is removed and it’s no longer a source of energy-rich terrestrial-derived prey items. Fish responses vary as well. We’ve seen increases and decreases in growth rates likely or as a result of elevated water temperatures, and cold limited systems such as those, some of those we have an interior Alaska, we might actually expect an increase in fish growth following a fire. A study in Idaho showed that trout in burn streams matured earlier and grew faster, which are characteristics of a colonizing, like history or population. They’re some interesting results there. At the population level, climate and fire interactions can drive population vulnerability of sensitive species. Here are the results from some recent work that we did on bull trout in central Washington. Bull trout require really cold water. We expect that the amount of available habitat, which is shown here on this graph, to shrink in the future relative to current conditions, or the status quo, likely or regardless of the climate scenario. The interaction here is that given the increases that we expect in the average fire size, it’s likely that in the future, the average fire sizes will actually be larger than the patches of bull trout habits, which means that if a fire happens within one of those habitat patches, it’s likely to burn the entire patch. These effects of fire on fish and population vulnerability can vary by life stage. It’s important to look across different life stages. Here’s another recent study that we did, again in central Washington, but now looking at Chinook salmon and breaking this out by eggs, juveniles, and adult life stages. What we found was that we predicted that the habitat quality, if you will, for eggs and fry and adults was lower post fire, but juvenile habitat quality actually increased significantly. Here’s a graphical representation of that. The colors on here correspond to the quality of juvenile habitat, with red colors being high quality. Here is pre-fire on the left, post- fire in the middle, and then I’ll draw your attention here to the right-hand side where is the change and habitat quality pre and post fire. We can see a lot of this orange on there, which is which is designating an increase in that habitat quality. Most of that is result of the increase of large wood to the streams, which is a really critical habitat component for juvenile Chinook salmon. Hopefully, I’ve shown you so far that fire effects on aquatic ecosystems are a complex web of interactions among fire and vegetation and geology and climate, etc. This is a well- illustrated figure from a recent review by Bixby and others. What we hope to do with this project and interior Alaska is to disentangle some of these, this complex web of responses of organisms to fire in interior Alaska. The project that we’re just getting started on has three main components, and with the overall goal of assessing ecosystem vulnerability to fire and climate change in the boreal forest, and identify factors that drive vulnerability to fire. For the first objective, we use a combination of field work and modeling to quantify patterns of fire impacts on aquatic ecosystems, then we’ll conduct a vulnerability assessment based on an integration of terrestrial and aquatic ecosystem models. Finally, we’ll use this information to work with managers to develop fire scenarios and decision support tools to assist with the aquatic species conservation in this fire prone or aquatic landscape in Interior Alaska. I’ll go through and briefly describe our plans for each of these objectives. Here’s a flow chart of the project. We’ll start over here on the left-hand side, the local scale. We’ll focus on bill work and some modeling results from that. First of all, I’ll tell you about the study area. The study area here is located in Interior Alaska. We’re focused on four river basins — the Chatanika, Chena, Salcha, and Goodpaster. These are all tributaries to the Tanana River, which is then a tributary to the Yukon. Fairbanks, where I’m at, is located as the green star here on the map. About 4,000 square kilometers had burned in this area since the early 1980s. Here, at the red, is the distribution of fire perimeters, and then in yellow is the distribution of Department of Defense lands. This is a large area. It’s about the size of the state of Pennsylvania. These basins support important fisheries, especially the Chena and the Salcha River basins are the first and second largest producers of Chinook salmon in the entire Yukon River Basin. All four rivers are important arctic grayling recreational fisheries. We’re going to focus on these two species for this project. This is another map of fires. I’ve broken it down by the burn history. We can see that there’s a mosaic of burned histories across the study area. Now, we’re going from 1950 to present. The older burns are here in green and the recent burns are in orange and brown. You can see there’s a lot more orange and brown on there than green. For the field part of it, we’re interested in contrasting sites that are located in recently burned areas versus historically burned areas versus controls or those that haven’t had fire since the 1950s. There’s no real controls in the boreal forest because all of these areas are in some state of post-fire succession. We’re going to use spatially balanced sampling design to select sites that are similar with respect to stream size, elevation, etc. The field piece will be limited to wadeable streams. This is for access reasons and logistics, plus we expect that the fire impacts are going to be easier to detect in the smaller streams. This part of field work at each one of our sites will install sensors that will continuously measure climate, stream flow, water temperature, and water chemistry. We’ve already started some of this last summer. Then, during the first summer of 2019, we’ll take a shotgun approach and sample a large set of sites across the four basins to look at broad patterns in fish community composition relative to those fire history classes. We’ll dial in and do some intensive sampling at a subset of sites within the treatment categories where we’ll collect a suite of the physical and biological data, which some of it is listed here. Through this field sampling where we’ve collected data necessary to parameterize two predictive models, we’re interested in looking at freshwater food webs to aquatic trophic productivity model. Also, we’ll use an individual-based model to look at fish biomass and population dynamics. We’ll also investigate these changes based on a range of observed conditions by fire history, and then incorporate our predicted climate impacts as well, which will result from the second objective, which was to integrate a suite of climate, terrestrial and aquatic ecosystem models, and then produce a vulnerability analysis based on those models. Our environmental and hydrologic models will be driven by this sophisticated climate models. We touched on this a little bit. We’ll use output from the weather research and forecasting model, which will give us historic and future precipitation, air temperature, and other climate variables. Using recently developed techniques, we’ll dynamically downscale these data from the coarse GCM scale all the way to 20, or even 4 kilometer spatial resolution for the study area. Steve suggested earlier we’re going to replace and model current and future fire and terrestrial vegetation and permafrost dynamics using two modeling frameworks, which have been developed in Alaska. These are part of the integrated ecosystem model. What we plan to do is to link the outputs from these models, the vegetation, fire, and permafrost changes to three aquatic ecosystem models. One of which covers watershed processes, NetMap, hydrology, the VIC model, and then stream temperature, which is called LST. I’ll talk a little bit about those right now. We’ll look at watershed processes using a program called NetMap. We’re working with Lee Benda at TerrainWorks to derive digital stream networks for our four study basins, based on a digital elevation model. This is really critical because interior Alaska currently lacks high- resolution digital representations of stream networks. We are having to build this from the bottom-up. What the NetMap model does is it produces these stream networks. Then, you’re able to predict in- stream, geomorphic attributes, such as channel type here in panel A. Also, conditions on hill slopes, such as soils, slope and aspect that control erosion and sediment delivery, which of course, are directly related to some of these fire impacts that I talked about earlier. We just received these data a couple of weeks ago, so they are hot off the presses. Just to show as an example of application here that we can do with this, we could, for example, identify reaches of stream that have high erosion potential. This might be a little bit hard to see here, but we can see these red reaches of stream are close to this hill slope, and we expect the erosion to be higher in this area. We can overlay that predicted vegetation and permafrost, etc. Look at how that varies throughout the study area. Basically, we have spatially explicit and continuous predictions across about 11,000 kilometers of streams that are located in our study area. Stream temperature is a critical control on fish growth and survival. We plan to use a model that we developed in Oregon, in Washington, to predict stream temperature across the study area, every one kilometer, at an eight-day time step. We’ll model stream temperatures based on the relationship between remotely- sensed land surface temperature, which is available from NASA’s MODIS platform, and then relate that to measured stream temperatures from our field study. Then, also previous data that we’ve collected over the past several years across these four basins. We’ll be able to predict future land surface temperature or emissivity, because it is one of the outputs of that work climate model. We’ve already implemented this to some degree in a Chena river basin, which is this figure down here below. This is a map of the gross potential for juvenile Chinook salmon. We’ve shown that the temperature model works well in interior Alaska. This is the result of this riverscape bioenergetics model that we’ve just wrapped up. It’s currently being considered for publication. We plan on expanding this model, reapplying it across the other three basins for this part of the study. Stream flow is also very important. It’s considered the master variable in stream ecosystems. We plan on modeling the spatial and temporal patterns of flow in the four basins using hydrologic model, which will be driven by output from the WRF climate model as well as vegetation conditions from the ALFRESCO model in permafrost conditions, as well from the DVM-DOS-TEM, the integrated ecosystem model. Then, this will all be routed through those stream networks that we’ve developed as part of that NetMap software. Each of our four basins has USGS flow gauges that we can use to calibrate these models. The basic framework for this is already in place. We’ve already done some preliminary flow regime classification for the Chena River, shown here in this figure on the bottom. We’ll expand this to the other basins and then incorporate this information from these other models. We’re also currently conducting a review of boreal stream flow regimes from which we can look at broad patterns of hydrologic variability and fire across the broader arctic boreal ecosystem. The second part of the second objective will use output from these integrated models to assess the vulnerability of Chinook salmon in the four basins to current and future fire dynamics in climate change. The plan right now is to use a Bayesian network approach to look at vulnerability. We chose Chinook salmon because they are species of concern. They’re very important for both ecologically, commercially, and for subsistence use in Alaska. They also spawn and rear over a large area within these basins, so they likely integrate the stressors within the population dynamics. Then we’ll produce some lifestage-specific estimates similar to what we did in Central Washington study. The Bayesian network approach is really appealing for a study like this because it allows us to track uncertainty throughout the models. It allows us to incorporate both quantitative and qualitative information. It’s easily updatable with new information. We can have spatially explicit predictions, and then we can incorporate multiple scenarios. Just as an example, here are some results from the study in Central Washington on bull trout. This is a similar approach here. Here’s a map of bull trout vulnerability to fire. The scale of these predictions is that of a one-kilometer stream reach. Red here are places that are more vulnerable, and green are places that are less vulnerable. Again, one of the key outputs from the Bayesian approach here, as far as I’m concerned, is that you can also track uncertainty. Here’s a map of the uncertainty in those vulnerability predictions. You can see some places that are predicted to have low vulnerability and low uncertainty in that classification, as well as some areas that are predicted to have low vulnerability but have a lot of uncertainty associated with those estimates. Those might be places that we want to update the model or look at individually a little closer. The last part of this project is to use a structured decision- making process to elicit input from fire managers in order to evaluate scenarios, the changing fire and climate dynamics on aquatic ecosystem. Fire management in Alaska differs a little bit from the lower 48. In Alaska, fire options are designated into four broad management options, everything from critical protection or full protection, which essentially put fires out, suppress fires, all the way to limit its protection, which is essentially to maintain the natural fire regime and minimize suppression. The other key piece of information here is that these options are set by the particular land management entity. In other words, managers select these options for their lands to reflect their objectives and their agency mission. Here’s a map of the current state of fire management options in Alaska. We can see here that about 65 percent of the state is in this limited protection and natural fire regime. Zoom in here to this area near Fairbanks and look a little bit closer. Previous work by Scott Rupp and his team developed a framework to evaluate the effects of changing fire management options on terrestrial ecosystems. For example, Fairbanks is located right about here, in the upper left-hand part of these graphs. The upper left blob here is the current management and then the lower right blob is new management option. What we could do here is look and see…For example, for these large tracts of DoD land outlined in yellow, what would be the implications of changing from a limited protection option up here on upper left to a full protection option here on the lower right? What would that look like? This is an example simulation that results for two climate scenarios. One that is a moderately warm and wetter feature and the other at some very warm and drier feature. The solid lines are that current fire management option and the dotted lines are the alternative fire management option. We can look at the predicted area burned under those two different options. They’re under the alternative options and these two different climate scenarios. The difference here between the solid and the dashed lines is the difference in area burned that would be predicted if you switched from, for example, a limited protection to a full protection across those two particular large tracks of lands. We’re going to utilize the same utility to offer our new project to interface with Department of Defense, state and federal fire managers, to explore fire management scenarios that may impact aquatic habitats. We’ll use the structured decision- making process shown here on the left to engage with and elicit input from managers with respect to scenario development. On the right are some examples of Web- based tools that were produced for this previous study on terrestrial ecosystems. They allow the user to map and do some data exploration on fire and climate impacts. This particular example is on terrestrial ecosystems. Our work will focus on aquatic ecosystems. The team that we’ve put together here has the capacity to build out these types of tools for managers and the public. That was all I have for you. There’s a lot of information and a lot of “This is what we’re going to do.” I do appreciate the opportunity to share this project with you. I’m very interested in any feedback that you could provide, either right now or later. My contact information is here. OK. Thank you, Stephen. Thank you, Jeff.

Leave a Reply

Your email address will not be published. Required fields are marked *