Fish that split their lives between fresh and salt water often face obstacles getting back and forth. Dams and roads fracture river networks and interfere with traditional migratory routes, sparking concerns about fish health and abundance, as well as biodiversity on a broader scale.
Efforts to restore fish passage are cropping up across the country, but these projects come with hefty price tags. In a new study, published June 3 in PLOS One , University of Washington researchers explore whether this money is being well spent by examining the process that determines which projects are prioritized.
The current standard, called score and rank, involves evaluating barriers one by one and assigning a score based on potential gains, such as habitat expansion. Top-ranking projects become leading candidates for funding, but score and rank systems don't always account for barriers in the full river context. High-scoring projects can yield stranded investments, where removing the barrier doesn't have the desired outcome because of other barriers downstream or immediately upstream.
"Ideally, barriers that are most downstream will score higher, because they need to come out before the fish can benefit from upstream restoration, but approaches to scoring vary, so this isn't always the outcome," said lead author Sunny Jardine , a UW associate professor of marine and environmental affairs.
As an alternative to score and rank, this study presents a mathematical computer program called optimization. Optimization synthesizes many inputs to make the most of a budget. It can serve as a performance indicator for other systems and highlight opportunities for improving an underperforming system.
"It's looking at a portfolio instead of going barrier by barrier. In doing so, you can explicitly account for watershed connectivity and evaluate the performance of score and rank," Jardine said.
As concerns about the health of rivers mounted in recent years, state and federal governments have allocated billions of dollars toward reconnecting them. Fragmentation is an established threat to biodiversity, and recent studies show that a vast majority of river length is not protected by conservation measures.
Washington state is in the midst of a court ordered multibillion dollar effort to remove barriers that block salmon and steelhead from swimming upstream to spawn. The statewide strategy combines score and rank with optimization in a hybrid approach. Similar projects elsewhere tend to use score and rank.
"I think people see optimization as a black box because it's not as obvious why a barrier rose to the top of the priority list," Jardine said. "With score and rank, they understand the scores and the process, but we don't really know what the outcome will be."
In this study, researchers use fish passage in Western Washington as a case study to compare score and rank to optimization. They show that score and rank performs decently well when the only goal is opening up as much habitat as possible, but adding other variables into the mix, such as habitat quality, compromises its performance.
While optimization has the capability to balance variables, it might not work for everyone. The program needs data to run and someone with a mathematical background to run it. Still, even small tweaks to the score and rank approach can produce results that rival optimization.
"Major change is hard, but minor changes may be enough," Jardine said.
Because these projects often represent the values of multiple stakeholders, it's important to include safeguards against stranded investments.
"You need to work from downstream up to make sure the success of a project isn't contingent upon other projects," Jardine said. "We're spending a lot of money on this, but the total cost of restoring all barriers is much higher than the budget, so it's really important that we make the most out of the financial resources that we have."
Additional co-authors include Logan Blair , a UW postdoctoral researcher in environmental and marine affairs; Catalina Burch , who completed this research as a UW master's student in environmental and marine Affairs; J Kahn, who completed this research as a UW master's student in quantitative ecology and resource management; Andrew Cooke, a UW research consultant in environmental and forest sciences, Luke Rogers , a UW research scientist in environmental and forest sciences; Mark Scheuerell , a UW associate professor of aquatic and fishery sciences and Robert Fonner , Daniel Holland , Connor Lewis-Smith , and Braeden Van Deynze of NOAA.
This study was funded by Washington Sea Grant and the Rae S. and Bell M. Shimada Endowed Faculty Fellowship in Memory of Warren S. Wooster.
