Skip to main content

National Center for Ecological Analysis and Synthesis

Jenny Seifert: Of all the fish people eat, salmon seem special. They’re not just celebrated meals, but also cultural icons. And they’re of incredible ecological importance in many places around the world.

In Alaska, salmon’s importance to nature and people is exceptional. For millennia, salmon have been a lifeblood of Alaskans, and the stories told about them are immeasurable.

There is one kind of story that is rarely told, however. It is about something that is literally measurable – Alaska’s data on salmon.

I’m Jenny Seifert, and this is a story of Alaska’s exceptional salmon data – and why they matter for the present and future of Alaska’s salmon and people.

Until recently, efforts to manage and protect Alaska’s salmon have had a bit of a data problem. It hasn’t been a matter of lacking data per se. In fact, there are loads of salmon data sitting on servers and in file cabinets across the state. But that was part of the problem. With so much data scattered about, they weren’t very accessible or useful to most people.

What’s more, when it comes to using data to make decisions about salmon, such as when people can harvest them, an important kind of data has been largely absent – and it’s not because it doesn’t exist. Those data are the knowledge and experiences of Native Alaskans.

Helping to fix this data problem is a project called the State of Alaska’s Salmon and People. It’s a huge endeavor, consisting of eight teams of experts who are compiling and analyzing as much of the state’s existing data on salmon as possible. They’re working to make those data more accessible and more inclusive of different ways of knowing. They’re also using them to draw a more complete picture of the state of Alaska’s salmon and the communities who rely on them.

And since this is a scientific project, and scientists love acronyms, it has an acronym: SASAP, which you’ll hear throughout this podcast.

But SASAP’s expert teams include more than just scientists. There are salmon managers from state and federal agencies, leaders from non-profits, and members of Native Alaskan communities.

In this episode of NCEAS Portraits, we’ll hear from members of two of the SASAP teams. And we’ll learn what their data can tell us about the changing relationships between Alaska’s salmon and people.

Part 1 is a story of an incredible amount of data that is helping scientists make sense of a peculiar trend – Alaska’s salmon have been shrinking.

Part 2 is a story of how data may – or may not be – a currency of empowerment in decision-making about salmon, particularly when it comes to conserving salmon and preserving an important way of life for many Native Alaskans, subsistence fishing.

Part 1: 14 Million Lines of Data

Seifert: An important dataset for managing salmon in Alaska is what’s called the ASL – that stands for age, sex, and length. These three simple measurements are collected from salmon caught by commercial and subsistence fishermen, and they are crucial for telling scientists and managers how salmon populations are doing.

Bert Lewis: And that’s part of the basic biological information that’s collected mostly from the commercial fishery tracking salmon populations across the state. And that started at earnest at statehood, but had already been in place before that, largely driven by the economic importance of salmon.

Seifert: That’s Bert Lewis. He’s works for the Alaska Department of Fish and Game as a regional supervisor in the division of commercial fisheries. His agency collects those salmon data. Bert also co-led one of SASAP’s working groups, which has been trying to understand a peculiar trend that scientists, managers, and fishermen have been observing in the past few decades – salmon have been getting smaller. 

Lewis: Our working group approached this issue by breaking it into kind of three subsections: consistency of the trends that we’re seeing in size and age across species, the causes – an investigation of what might be driving this, ocean conditions, climate change, freshwater environments, predation, competition with other salmon – and then the consequences. What are the economic, ecological, and social consequences of the trends that we’re seeing. And that approach allows a determination of what’s going on, why, and then what the impacts might be.

Seifert: Understanding the consistency, causes, and consequences of these biological changes helps managers figure out how to adapt their plans so they can sustain salmon into the future. To gain that understanding, Bert and his working group collected all of the ASL data from every regional and field office around the state.

Lewis: The legacy of starting all these data collection projects across a state as big as Alaska meant that each region and even each field office – if you think about Cordova, or Nome, or Sand Point – was collecting their own data, and it was archived in file cabinets filled to the day. And getting it all electronic format and assembled into a single location for the first time was something that hadn’t been done.

Seifert: The resulting dataset produced an astonishing amount of data – 14 million lines of it.

I spoke with Bert and another member of his working group, Krista Oke, about this impressive dataset. Krista has been a postdoctoral researcher on the project and is currently based at the University of Alaska Fairbanks. In the following conversation, Bert and Krista explain why this dataset is important, what they are finding out about this trend of shrinking salmon, and what it all could mean for salmon – and Alaska – moving forward.

Seifert: Why is it valuable to have all those data now compiled in one place?

Lewis: It lets you track trends on a broader scale; whereas each field office might notice and talk about internally, ‘hey it looks like the Sockeye have gotten smaller, and we’re missing our older age classes.’ Now you’re able to compare and contrast across regions. And these data are used to, what we call, populate our brood tables, which is an assessment of the number of fish that we turn from each spawner. And that allows us to forecast how many fish are coming back in the next season, which is important for management of the commercial industry, to understand what resources they need to be prepared to bring. And it also is the foundation of our escapement goals. And that is the goals of how many fish we need from the spawning population for sustainable management. So, the age and size data, along with the abundance, is all used as the foundations of our sustainable management. And Alaska has pioneered what we call in-season escapement, goal-based management. And if you look at fisheries across the globe, many are overfished and in depressed status, and they’re managed by setting a goal of harvest before the season and going out and taking that regardless of what’s going on. We track the salmon returns in season, and adjust our management of harvest according to the abundance and timing of fish as they come back. And that’s unique to the state of Alaska and really is well adaptive for the salmon life history, the way they come back and home to their natal systems.

Seifert: This question is for you Krista, you’ve been very involved in the data analysis for your group. What has it been like to work with all this data and synthesize them? Did you encounter any data mysteries or oddities or surprises along the way?

Krista Oke: I think it’s been really amazing to work with this data. You mentioned earlier, we’re talking about 14 million rows of data. That’s a dataset that you can’t just open in Excel. You can’t just load it into R and look at it. And so, it’s been really challenging to learn the process of how do you quality check, how do you look for typos and everything like that in a dataset like this? In terms of surprises, I would say that there’ve been a few. The main one for me, I think, was the relative lack of data on pink salmon. SO pink salmon are a really abundant species but there just hasn’t been as much data, at least for size and age data, collected on them. And that was really eye opening to me. And in the end, we actually had to exclude pink salmon from our study, just because we had so few data. And then I think another one that was really surprising to me was that, you’d think with 14 million rows of data, oh we can run any kind of analysis we could dream of on this dataset. And what we found is that, in a lot of situations, we are actually running into a situation where we still don’t have enough data, particularly for analyses that require really long time series or comparing among a number of different populations. And so that’s been really surprising to me that, even with 14 million rows of data, you can still not have enough data to run the analysis that you wanted to.

Seifert: I know that you’re stilling preparing publication manuscripts, but do you have any major findings or insights that you’re comfortable sharing now?

Oke: One of the things that we were really interested in for our study was that there are two different ways that a salmon population on average can get smaller. First, it could be that, on average, the growth rate has slowed down, such that each fish is smaller at the same age. Or the second option is that you might see shifting age structure, such that fish on average are coming back from sea at an earlier age. And so, one of the things we tried to do was to tease apart whether it was shifting age structure or shifting growth rate. And what we found was that, overwhelmingly, it seems to be shifting age structure. So, if fish are getting smaller, it’s mostly because those fish are actually younger, and in most cases, it seems that shifting age structure is accounting for somewhere between 60 to 80 percent of the size change that’s happened.

Lewis: And we knew that salmon were getting smaller, but the extent and consistency between species is probably the most interesting and formative finding that we’ve had so far. And then determining what contributed to that decline, whether its size at age, or age, and finding that it was age, with fish coming back younger as the primary cause behind this, those two things are the biggest findings that are going to drive management decisions, future research, going into the next steps.

Oke: Also, I wanted to mention that we’ve also been trying to determine the causes or the drivers of these size changes, and we’re really right in the middle of those analyses right now. But what’s very clear is that we aren’t finding any single driver of size change, and instead what we’re seeing is multiple different factors contributing to size changes. So even with all this data from across the state, we don’t find a smoking gun, and it seems instead that it’s lots of different factors that are contributing small amounts to overall size change.

Seifert: What are some of those factors?

Oke: So, the ones that we were really interested in were fisheries-induced evolution due to selective fishing gear.

Seifert: What does that mean exactly?

Oke: So often, one of the kind of consequences that you see of intensive fishing is that, to mitigate the risk of being caught by a fishery, fish will often evolve an earlier maturation age, and often also smaller size at maturation, which allows them to spend less time in the ocean if the ocean is really risky, and instead to reproduce as fast as they can to try to make sure that they reproduce before they get caught by the fishery. Some of the kind of classic signs of fisheries-induced evolution are smaller size and younger age. Some of the other main potential drivers include climate change. So with warming, we’ve seen body size decreasing across a number of different taxa, including insects, other fish, mammals. Also, predation from rebounding marine mammal populations, in particular orcas. And finally, we’re also really interested in competition among salmon species. In the past few years, salmon abundance has been higher than they ever have been on record, and so there’s been a lot of attention recently to looking at whether salmon might be competing at sea.

Seifert: How do you hope your data and findings will help salmon management and decision-making in Alaska?

Lewis: Well, the fact that we’re able to look on it on a broader scale and compare and contrast patterns across the state of Alaska – because, if you think about starting down in southeast Alaska near Canada, and then out the Aleutian peninsula, and then up into the Bering Sea, there’s so much variation and geographic separation between those. Being able to look at this as a whole is going to help us understand implications for management. And then some of the consequences about how is it going to affect in-season escapement goal-based management, as I touched on earlier. And is it going to affect our escapement goal analysis and the way they’re calculated. That’s the foundation of our salmon management. So, it’s going to allow us to consider the broader consequences of these trends and the analysis of what the potential drivers are, and it’s going to allow us to maintain a management approach that could keep populations sustainable into the future.

Seifert: Why do you think a synthesis project like SASAP is valuable to salmon science and management?

Lewis: So, the SASAP process of bringing in an interdisciplinary group with new perspectives and then the whole approach of having – I call this a salmon think tank – of having this group get together with the data and be sequestered for five days of really intensive conversation, and then looking at data and debating, sometimes heated debates about what they mean, as part of the scientific process, really allows the exploration of all the different aspects and directions it might go. And that’s super productive, and as a state agency, we don’t have the time and capacity to do that type of detailed work, especially the sequestration of this group away from their offices that really allows a focus that anyone who has a job knows it’s very hard to achieve in your normal day-to-day process.

Oke: I think there’s a lot of value in comparing trends among regions and even among species. So there has been a lot of work done on this question in the past, but I think being able to take this synthesis approach and really compare across species has led to a lot of realizations that we might not have gained just by looking at individual species. And so, I think we gain a lot just by being able to compare at a statewide level. And from the perspective of just future science, I think this is a really valuable kind of snapshot of what is happening to size change right now, or up to right now. And having this dataset and this study available I think is going to kind of spur a lot of new projects looking more in-depth into this question.

Seifert: My last question is what’s one big takeaway you’ve each learned as a result of your involvement in SASAP, especially as it might relate to data synthesis to inform salmon decision making?

Oke: I think for me the biggest takeaway is just how complex this issue is. I think it’s really important to do these synthesis type projects, where we can really take a step back and go beyond comparing just what’s happening in each individual location and look across the state and compare trends. But, for me, one of the main takeaways is just it’s so complex.

Seifert: How about you Bert?

Lewis: Echoing what Krista had said is that, even with 14 million lines of data being collected since 1960 and earlier, that the complexity of these systems, of natural systems, especially something like the ocean, that even with bringing all these resources to bear and some very bright minds, the fact that we really haven’t found a single driver responsible for this, just brings home again the complexity of these natural systems and our inability to even understand, much less manipulate as desired any of this is tremendously challenging. And it just introduces uncertainty and suggests that, whenever we’re managing and working in natural systems like this, erring on the side of conservative management and using a precautionary approach are the best way to move forward.

Part 2: Data as a Currency of Empowerment

Seifert: The data Bert and Krista are working with are primarily from commercial fisheries. But there’s another important source of data that helps with the in-season management of Alaska’s salmon – data related to subsistence fishing. Our next story is about how these data may or may not be a currency of empowerment in salmon decision-making.

An area of Alaska with a long and rich tradition in subsistence fishing is the Kuskokwim River Basin. The river stretches over 700 miles through southwest Alaska, and there 33 rural communities living along it, which are inhabited primarily by Native Alaskan tribes.

The Kuskokwim is host to the state’s largest subsistence harvest of Chinook salmon, also known as the King salmon, a species that is prized for its size and nutritional content. And it is especially important to the cultures of the Kuskokwim’s communities.

Nick Kameroff is an Alaska Native and comes from Aniak, which is one of those communities.

Nick Kameroff: My relationship with salmon is my lifestyle I would say. Ever since I was born, I remember, once I was able to walk, carry my body on my two legs and walk over to the edge of the bank and sit there with my grandpa and watch, I used to ask him what we were doing. And he said, we’re fishing for the winter, this is our food for the winter, you have to learn how to do this. So, my relationship with salmon is like anything else throughout our wonderful world up here we live in, is that, I’m harvester, a gatherer, subsistence fisherman. And a provider for my family as well as several other families when needed.

Seifert: In addition to being a fisherman, Nick is an in-season manager for the local Intertribal Fish Commission. During fishing season, when the salmon make their way from the Bering Sea and up the Kuskokwim to spawn, he works alongside state and federal fishery managers, using data to determine when and for how long communities can fish, and how many salmon they can harvest.

Kameroff: And it’s our job to try to make sure everybody along the whole drainage gets a chance at the resources.

Seifert: One of main concerns for Nick and his fellow managers at the moment is helping the river’s King salmon recover from a population crash that happened in 2013. The crash has been deeply troubling to the local communities, and data is playing an important role in guiding decisions about conserving the King.

Kameroff: This in-season monitoring lets us know what’s in the river, where it is, how many, how much of them. And this data is turned around, it’s gathered relatively quick within two or three days we have it turned around, sometimes within a day and a half. It’s turned around for us to look over and see what’s actually there. And we have to be really careful of overharvesting because like I said we’re rebuilding the King salmon stock. I feel that the in-season harvest monitoring is a really great asset for us to, you know, make sure that we make wise choices for the stakeholders along the river.

Seifert: An ultimate goal of our second SASAP working group is to support and empower Nick and his fellow managers in making these wise choices, especially in conserving the King and enabling equitable access to salmon as a resource. Nick was part of this working group, along with several other tribal community members from the Kuskokwim River Basin – as well as several scientists from universities in Alaska and beyond.

Mike Jones was a lead researcher for the group. He’s a fish biologist and professor at Michigan State University. As a native of Vancouver, British Columbia, salmon have long been part of his life too.

Mike Jones: I had the good fortune of being interested in biology and had a father that worked in the salmon fishing industry. And spending two summers in the central coast of British Columbia, working as a deck hand and seeing the salmon commercial fishery, the salmon commercial fishery operate, I just fell in love with it. And so, it became what I focused on as a scientist, as I studied biology. I said, well I want to, I like biology, I want to study fish, why not salmon?

Seifert: Unlike the other SASAP working groups, Nike and Mike’s team didn’t really do a data synthesis, meaning they didn’t combine a bunch of different datasets to run big scientific analyses about salmon. Instead, they were focused on data needs – or, to put it in a more positive frame, how can data help empower decision-making?

First, they wanted to know what data would be the most helpful to salmon managers in determining the subsistence harvest. Second, they also wanted to know how can local communities help managers collect important data while also feeling their voices are included in management decisions.

Answering that first question, the one about managers’ biggest data need, gets a little technical. It involves mathematical models and a concept called the value of information. I’ll let Mike explain.

Jones: Models represent how describe the workings of the world in a synthetic way. Our group was started from a point of view where we had models that we had built through earlier research, and we wanted to use those models as what I would call synthetic tools to ask this value of information question. In other words, to ask where do we get the greatest bang for our buck in terms of reducing uncertainty, and then to what extent does that sweet spot intersect with opportunities for community-based monitoring?

Seifert: In any decision related to natural resource management, there is going to be some uncertainty – how well can we predict when the salmon are going to run, or do we have enough data to understand the true size of the salmon population? Models typically help reduce this uncertainty by pulling together a bunch of different data to mimic natural processes and estimate the outcomes.

Jones: The sort of scientific basis of the work that we did was an idea known as value of information. And the concept is basically that you use the models that are the basis for management decisions for fisheries in general, models that forecast run size and so on and so forth. You use those models to ask questions along the lines of ‘what if we were more certain about some aspect of the model, how much better off would we be in terms of the quality of the decisions that we could make?’ The quality of the decision measured in terms of, say, reducing the risk of failing to meet your escapement goal or something like that.

Seifert: If you recall, escapement is fisheries speak for the number of salmon that escape the fishing net and make it to their spawning grounds to breed.

Jones: So, by using this sort of method of asking how much would we gain by reducing some area of uncertainty, we can kind of objectively define what information is most valuable. One of the things that we found in this working group project is that, in fact, the most important information for managers to be able to make the best possible choices about fishing in season, is actually the quality of the pre-season forecasts – the prediction of how many fish are going to come back at the beginning, sort of before you even start fishing. The degree to which we can accurately forecast the run size really has a big effect on how closely we can manage the fishery to meet the goals of allowing enough fish onto the spawning grounds, the escapement goal, and yet still allowing as much harvest as is safe to take.

Seifert: By reducing the uncertainty around how many salmon could come back to the river in the coming season, pre-season forecasts will hopefully be empowering tools for managers as they decide when and how much salmon the communities can harvest sustainably from the Kuskokwim. But can data also empower the communities who depend on the subsistence harvest?

That brings us to this working group’s second big question, which focused on a type of citizen science program that is often seen as a way to empower local communities: community-based monitoring programs. These are run by local tribal councils and fishing associations, who enlist tribal community members to collect data about salmon and subsistence fishermen, which then contribute to decisions about the subsistence harvest. But are these programs actually empowering to the tribes? In other words, by giving them the opportunity to provide data and input to managers, do they feel they get a voice in salmon management in return?

Helping to answer this question is Janessa Esquible. She is a fisheries biologist and runs the community-based monitoring program for the Orutsararmiut Native Council, the largest tribe in western Alaska, which is based in Bethel, the largest village in the Kuskokwim basin.

Janessa Esquible: And we at ONC are one example of a community-based monitoring program, where the tribe virtually kind of executes the project. We have a local entity executing the project on the ground. We do partner with Alaska Department of Fish and Game, but it is very community oriented and community driven.

Seifert: The people who collect the data, or the monitors, play many roles in their communities – they’re youth leaders, mothers, fathers, fishers, tribal leaders. And they are becoming increasingly important to salmon management in the state as more boots-on-the-ground collecting data, which is especially important in hard-to-reach places like the Kuskokwim river basin.

Esquible: Without having a commercial fishery operating here, managers have no idea what the harvest looks like during the run, during the fishery operating, and it’s important for them to get an idea of how well are our fishers doing out there, are they seeing abnormalities, what concerns do they have about the way the fishery is operating right now.

Seifert: The monitors collect a variety of data from fishermen during the harvest season.

Esquible: We have the harvest data and that includes information such as the gear type that they use while they’re fishing, how long their net was in the water actively fishing, how many salmon species they caught, and progress, how close are the fishers to achieving their subsistence salmon needs.

Seifert: They also collect basic biological data from fish caught – that’s the Age, Sex, and Length data that you might recall from Part 1.

Esquible: One other really important component of the survey instrument, and that’s to document any comments or concerns that fishers may have about the fishery. They might be seeing abnormal lesions on the fish, they might be seeing smaller fish than they’re used to seeing. And they’ll also share any comments or concerns that they have about the way the fishery is being managed.

Seifert: The tribal council will then relay all these data – from fish counts to concerns – to the management bodies to help them determine when to open or close a harvest, and how much salmon people can take.

Esquible: And that part is really important, because many of the fishers either can’t attend these meetings, can’t call in from fish camp, or they just don’t really want to. They don’t feel comfortable calling in or coming to the meetings in person and sharing their concerns. So we kind of serve as a conduit for relaying that information.

Seifert: This role, as conduits of information between the tribes and the managers, is partly why many people see community-based monitoring programs as a way to give the tribes a voice in decision-making. But is that actually the case in the Kuskokwim? Janessa explains from the perspective of the tribal council she represents.

Esquible: It is a way of empowering the tribe, because the tribe is hiring local technicians. They’re in constant communications with local fishers throughout May, June, and July. They’re gathering information that’s not only important for managers, but also, like I mentioned earlier, comments and concerns that fishers want us to give and relay back to the Fish and Game managers, researchers, and other community members.

Seifert: But they also wanted to know how the fishermen felt about the program.

Esquible: And do they feel their voice is heard by being asked every week to provide us with this information? What we learned was that, you know it was about 50/50. We had about half of the respondents that felt they benefitted from the program. It was interesting, though, because the harvest data component wasn’t really cited as a way of them benefitting. What they did tell us was that they liked the ASL sampling program because they can get paid $5 for fish that they sample. So that was a way to kind of earn money maybe for their gas while they’re out fishing or whatever. And then they talked about some of our other programs within this project. And that includes things that aren’t really a part of the community-based monitoring, in collecting harvest and biological data.

Seifert: Those non-data benefits included things like the calendars that the program gives to fishermen to help them keep track of their harvests, or the fish they distribute to elders, or their tips on when there will be another opener.

Esquible: But for those that did not feel their voice was being heard, the theme was this distrust. They felt there was a disconnect between the data we were collecting from them and the management action. So, for instance, if we have folks that are telling us, let us fish early, you know, the rainy weather is here. We need to fish when the weather is dry, with good wind to help with their preservation of their fish. And so, when they’re sharing these comments with us, and then the fishery remains closed, to them that’s a disconnect. They feel that their voices aren’t heard when there’s not an opener even though they’ve told us, hey we want the fishery to be open. And I think we’ll continue asking these questions, because we can’t assume that community participants are being empowered through this tool.

Seifert: Janessa explained these mixed responses can help them improve their program. But it also points to an important consideration when designing community-based monitoring programs with the intent to empower.

Esquible: And so, unless the community-based monitoring program was indigenous driven, or where the research was directed by the tribe or community itself, we can’t really make these assumptions. I think having communities involved in the programs is definitely critical for long-term sustainability, especially in rural Alaska or off-the-road-system communities that are more difficult to access. And I think if other regions begin to employ these CBM programs, I hope that they engage communities from step one to ensure that they’re involved in identifying critical information needs, in addition to those identified by scientists. And I just think it’s really important for locals to have a full sense of ownership of the project, and this would really contribute to the overall success and implementation of such a project.

Seifert: So, in thinking of data as a currency, as the generation and exchange of something of value to decision-making about salmon, a sense of ownership over what you are offering and what you get as a result is critical to feeling empowered in that process. But equally important in that exchange are trust, respect, and a willingness to listen. Here’s Mike again.

Jones: I would say this is true of the SASAP working group for me, as well as other things that I’ve done, is that it’s just enormously helpful as a scientist, when you’re working on real world problems, it’s just enormously helpful to be able to have a conversation with the stakeholders who are actually depending on the resource that you’re trying to apply your science to. I think for all of us, this experience has just, it’s created a broader perspective. A perspective that recognizes the points of view of traditional users of salmon and the cultural importance of salmon, at the same time as a broadening of perspectives of others who are very intimately familiar with that and develop an understanding of how western science can help to inform thinking.

Seifert: And for people like Nick, the reasons for empowering his community in decisions about salmon transcend the data.

Kameroff: Our people have lived off of these resources for thousands of years, but unless like they’re doing now, realize when we have issues like conservation, that if we don’t take action now, then nobody will have a chance at that resource in the future. I always tell our people in our meetings, when we have a huge Intertribal Fish Commission gathering once a year, you’d rather have your children be able to work – your future children, your children’s children and grandchildren, great grandchildren and so forth – you’d rather have them have the opportunity to look at, touch, work, cut, and eat that fish versus go to a museum and wonder what they looked like, or how they felt, how they tasted, and everything else. Because that’s our food security we’re talking about. And I think as time goes, we’re going to be the go-to people for this various management of resources.