A new tool weighs the multifaceted factors involved in efforts to improve habitat

From the Summer 2025 Sea Star

By Brian McGreal, WSG Science Communications Fellow

In 2018, in a case brought forth by 21 Northwest Tribes, the Supreme Court ruled that Washington State has a duty to protect tribal treaty rights by preserving fish runs. The main focus of this case was replacing culverts — infrastructure that allows streams to pass underneath a road or path, and are also commonplace barriers to migrating salmon.

Through a Washington Sea Grant-funded research project, Sunny Jardine, an economist at the University of Washington (UW) School of Marine and Environmental Affairs, along with collaborators Mark Scheuerell from UW School of Aquatic and Fishery Sciences, and Robby Fonner and Dan Holland from NOAA Northwest Fisheries Science Center, spent two years evaluating how to go about improving salmon habitat through replacing culverts. Given the high level of human activity along Puget Sound, the cost of effective salmon recovery efforts are the highest they’ve ever been, while the likelihood of success is low. Under such conditions, maximizing return on investment is essential. With a new webtool, Jardine and team have provided a way to optimize the decisions behind making culvert-ridden landscapes more accessible to salmon.

For decades, populations of Chinook and sockeye salmon and steelhead trout have been listed as threatened or endangered in western Washington. Salmon recovery efforts in the state are multifaceted, with an emphasis placed on salmon fisheries management and hatchery reform. Perhaps most critical, is the effort to restore salmon habitat that has been lost or degraded due to human activities.

Habitat degradation takes a variety of forms, from stormwater runoff carrying pollutants into spawning streams, to humans clearing and developing floodplains, to lowered flows in streams due to water diversions for municipal and agricultural uses. Physical barriers also prevent salmon from accessing otherwise healthy habitats. Dams are likely the most well-known of these barriers, but culverts are perhaps more detrimental to salmon in western Washington.

In 2024, the Washington State Department of Transportation (WSDOT) documented over 2,000 culverts acting as barriers to fish passage. Culverts take different shapes and sizes, from box culverts which are in essence small bridges, to concrete pipes allowing a stream to pass under a road instead of over it. Puget Sound residents likely walk or drive over them every day without realizing. Culverts can impede salmon passage due to being damaged, by being blocked with debris, or, in the case of narrow culverts, by increasing flow rates beyond the ability of the fish to swim against.

After the Supreme Court case, the state began funding efforts to restore salmon passage by improving culverts. WSDOT completed 32 barrier renovation projects in 2023, opening over 65 miles of potential upstream habitat to spawning salmon. But with thousands of culverts creating barriers to fish passage and limited funds available to address them, the question becomes how best to target habitat recovery efforts on a statewide scale.

A question to which Dr. Jardine was paying close attention. Jardine’s background studying optimization problems led her to consider how choices were being made to reclaim the most salmon spawning habitat possible from culvert restoration projects. “In order to have any chance we really have to make sure the conservation actions that we’re taking are the best we can do for the amount of resources that we have,” Jardine says. How are the best possible decisions made under restricted conditions?

We encounter optimization problems every day, without necessarily realizing it. You walk into a grocery store with a limited budget, knowing you need to meet your baseline nutritional needs for the week, but also hoping to maximize the satisfaction you get from the foods you eat based on your tastes and preferences. In this type of experience, an agent (you) tries to minimize or maximize some outcome (satisfaction from the foods you eat) while adhering to a set of constraints (your budget and nutritional needs). In the case of salmon passage restoration in western Washington, the underlying principles are the same, but the problem is much more complicated. Often there are multiple culverts in a stream system, with varying levels of passability for fish and differing costs involved in restoration efforts. Salmon cannot gain access to potential habitat upstream of a barrier unless all those barriers downstream have also been improved. With thousands of fish passage barriers situated along various stream networks, finding the set of restoration projects that will maximize habitat gains for a given budget is no simple task.

Jardine’s work began by matching the locations of barriers from the state barrier inventory with maps of stream systems in western Washington. Post-doctoral researcher Braden van Dynes also developed a cost model for barrier restoration to be applied to the inventory. With these components, Jardine constructed Upstream, an open source web app that policymakers and managers may use to determine the optimal set of barriers to target with restoration projects based on their available budget. This tool is designed to be applicable beyond salmon in western Washington: as long as a user can meet the app’s data input needs, it can be applied to the restoration of migratory fish habitat anywhere in the world.

Jardine is interested in comparing the optimal choices found through her work to those choices arrived at from pre-existing approaches. Essentially, how close to optimal have policymakers and managers gotten with their culvert restoration decisions in the past? Where have prior decision-making processes succeeded and where have they struggled? Thus far, this work has found that existing frameworks perform fairly well against the optimal benchmarks. These are typically less computationally intensive and data hungry than large scale optimization models, making them cheaper and easier to implement. However, this evaluation is still ongoing, with the hope that the optimization tool will be able to shine a light on the tradeoffs between seeking optimal solutions or employing commonplace frameworks for prioritization.

As this project moves forward, an important next step will be to consider variability in the quality of habitat being made available. Jardine is working to incorporate stream temperature, an important element of habitat suitability, into the model. Integrating projections of future stream temperatures will also allow for choices to be made based on what will likely be suitable salmon habitat not only today, but also decades from now.

Given the opportunity to meaningfully improve the health of Washington’s salmon populations, it is of vital importance that the best possible choices are made. The work of Jardine and her collaborators has provided policymakers and managers with an extremely useful tool to enable optimal decision making. With the future of salmon at stake, the more barriers we can move to effective decision making, the better.

Visit the Upstream app.

###

Washington Sea Grant, based at the University of Washington, helps people and marine life thrive through research, technical expertise and education supporting the responsible use and conservation of coastal ecosystems. Washington Sea Grant is one of 34 Sea Grant programs supported by the National Oceanic and Atmospheric Administration in coastal and Great Lakes states that encourage the wise stewardship of our marine resources through research, education, outreach and technology transfer.

wsg.uw.edu

Join the conversation: instagram.com/waseagrant and Facebook.com/WaSeaGrant.