Drought and Fish in the Willamette Valley

Imagine a very bad drought year hitting Oregon’s Willamette River Basin. The drought leads to low river and stream flows that impact the basin’s endangered fish species. Now imagine strategies that could help reduce these impacts.

This is the subject of a study published last July in the journal Nature Sustainability. The study, led by Oregon State University researcher William Jaeger and including CIRC researcher and University of Oregon law professor Adell Amos, presupposes a severe drought year in the Willamette River Basin in which the basin’s total stream and river flow is down by 55%.

Here’s the drought scenario in more detail.

Because low flows generally mean warmer waters, these low flows on the Willamette River and its tributaries present a major problem for the basin’s endangered and threatened fish species, including Chinook salmon and steelhead.

With waters especially low in the basin’s rivers and streams, several fish species could overheat and die, as millions of steelhead did throughout the Columbia River Basin, including the Willamette River, during the 2015 drought. (Here we should note that the analysis performed by Jaeger and colleagues did not look directly at stream temperatures, but rather took warm stream temperatures to be a kind of given that would result from low stream flows.)

Under the Endangered Species Act (ESA), federal and state agencies are required to meet certain minimum instream flows in the Willamette and its tributaries in order to protect endangered and threatened fish species. This means that the Army Corps of Engineers, which operates the dams on the Willamette River’s tributaries, must release enough water to meet a legally required minimum amount of water flowing in sections of the river and its tributaries.

However, in Jaeger and colleagues’ drought scenario, the ESA-mandated minimum flows cannot be met. There simply isn’t enough water in the reservoirs.

But what if we could turn back the calendar and anticipate the drought coming before it happened? How would we reset local policies to meet the ESA mandate? What policies would we change and how? These are the questions Jaeger and colleagues explore at length. The answers they came up with are, as the study’s authors put it, “complex.”

Let’s restate the low water/hot fish problem in more detail, and add on a few more complications for good measure.

The calendar period to meet the ESA requirements runs roughly from the end of March to beginning of November, which corresponds with the upstream spawning voyages of adult fish and the downstream journeys of juvenile fish. However, this calendar period is also the dry time climatically for Oregon, which, like much of the Pacific Northwest, receives the majority of its precipitation during the cool winter months. And there’s another complication: the region’s dry summer months are also the peak period for demands on water by the basin’s human residents.

If you’ve ever spent any amount of time in the Willamette River Valley, it’s apparent just by looking that the valley contains some of Oregon’s most productive farmlands. Many farmers in the valley rely on water from the Willamette River to irrigate their crops during the dry summer months. And then there’s the basin’s cities.

The Willamette River Valley is Oregon’s most populous region and includes the Portland metro area, Salem, Corvallis, and Eugene-Springfield. Add it all up and what you get is multiple peak demands on Willamette River water happening at roughly the same time. This calendar period of peak demands also roughly overlaps with the same period that fish need the water in their streams. That’s a problem.

Jaeger and colleagues identify three sectors where policies could be altered to get that needed water back in the river and its tributaries. They are farms, cities, and the reservoirs themselves. The sectors fall into two big categories: supply (the reservoirs) and demand (farms and cities). Let’s put aside reservoirs for now and focus on the demand side of the problem, those farms and cities.

To get farmers and city dwellers to conserve water, Jaeger and colleagues propose a classic mechanism to limit demand: they make the water available to cities and farms more expensive. (Though they note that other regulatory approaches could serve the same end.) For city dwellers, this entails raising the cost of water by 40–160%. For farmers, the researchers propose imposing a fee on farmers of $10–$40 per acre of water.

Now, let’s consider the supply side.

The Willamette River has 13 tributaries with 13 dams on them. The dams were built primarily to reduce the risk of flooding. They are now operated for flood mitigation, summer water storage (for uses including crop irrigation), and to manage flow for multiple purposes, including meeting the minimum flow requirements for fish. In practice this has meant that dam operators play a kind of game each spring (when snow melts and rains are heavy) of trying to the fill their reservoirs just enough to have sufficient water for summer use but not so much that they risk flooding everyone downstream if, for instance, a huge storm were to suddenly dump too much water on the reservoirs. This process is driven by what are called rule curves, which are essentially operating procedures that guide how the dam operators fill their reservoirs gradually from February to June. The idea is to find the Goldilocks amount of stored water while still being able to mitigate the risk of flooding.

Here’s where Jaeger and colleagues work gets really interesting and wonderfully wonky. A little backstory is helpful.

The Nature Sustainability paper comes out of the Willamette Water 2100 research project (WW2100). A National Science Foundation-funded effort that CIRC participated in, WW2100 is an ongoing effort to anticipate how water availability is likely to change in the Willamette River Basin as drivers, including climate change and population growth, affect both the supply and demand for water in the basin to the year 2100. To figure this out, WW2100 researchers modeled as much of the Willamette River Basin as they could. This meant tracking the complex interactions between the basin’s natural and human systems and then modeling that interaction in a computer. In this case the computer model was Willamette Envision, which was based on the ENVISION software developed by former CIRC researcher John Bolte.

WW2100’s modeling effort involved figuring out the physical climate and hydrology of the basin; as well as what the demands were on the basin’s water supply from farms and cities, the rule curves at the basin’s reservoirs, the ESA requirements for fish, and the water rights law that determines who gets water if there’s a drought.

But where the project really excelled is in how it manages to make all this information specific to the actual geography of the Willamette River Basin by putting that information on the map, so to speak.

Essentially, the WW2100 researchers built a spatial, two-dimensional model of the basin that includes the daily flows of water, the location of the basin’s reservoirs, the basin’s cities, the basin’s farmlands, the basin’s water rights (for farms and cities, both surface water and groundwater, and where the water can be drawn from), and the sections of Willamette and its tributaries that must meet ESA-mandated minimum flow requirements for fish.

Having these nitty gritty spatial details has been key to the project’s unique findings. For instance, in a previous paper, the project’s researchers examined just how many farms might benefit from irrigation in the future. The researchers determined that while the number of farms that could benefit from irrigation was high at first glance, given the distances from water diversion points to farms, the cost of transporting water to those farmlands was too costly to make it profitable. So much so that the actual number of farms that could benefit from irrigation was much lower than it at first appeared. Something analogous to this occurs in the thought experiment posed in Nature Sustainability. Here the story is one of both space and timing.

As noted above, to determine if ESA requirements could be met in their terrible, horrible, no good, very bad drought year scenario, Jaeger and colleagues looked at the combined savings (in terms of conserved water) of reducing demand for water on the part of farms and cities and they looked at increasing the supply of water by releasing water from the basin’s reservoirs.

Like farms benefiting from irrigation in their previous study, it turns out that the amount of water that could be conserved looks large at first, until you considered the actual timing and location of when and where water demand could be reduced. When geography and timing are considered, savings from both farms and cities are much less than you might expect because the reduced demands from cities and farms just doesn’t line up in the locations where and at the times when the streamflow shortfalls occur.

Let’s start by considering the location of the basin’s cities in relationship to the location of those ESA requirements. Most cities are located along the Willamette River itself rather than one of its upstream tributaries. However— because water tends to flow downhill—the location of the majority of the ESA low flow sites are upstream from these major cities. (The notable exception here is Eugene-Springfield.) In the researcher’s scenario, saving water downstream—which raising water rates in the cites would do—did not put water back upstream where it was needed for fish.

That’s space. There is also a timing piece to city savings as well. The calendar period of the cities’ largest demand for water also tends to occur after the largest shortfalls in streamflows. Again there is a mismatch, and only moderate savings occur.

Agricultural demand tells a similar story of space and timing not matching up. Jaeger and colleagues found that only 40% of irrigated farmland was upstream from where ESA-mandated low waters occurred in the drought scenario. In other words, over half of the irrigated land was downstream from those mandated spots, where saving didn’t really matter. But the real kicker had to do with calendar timing. The researchers found that while 60% of ESA-related water shortages is expected to occur by May, only 4% of irrigation is diverted by that date, meaning most farms were not competing with fish at the time water is needed to offset low flows.

So much for the demand side. What about the supply side, the reservoirs? Here the story is quite different.

The combined water conservation strategies of limiting demand from farms and cities and increasing supply at reservoirs was able to mitigate only 24% of the ESA shortfall. Of this available water, a whopping 81% was attributable to changes in the amount of water stored at the 13 reservoirs, while reduced agricultural demand accounted for just 16% of savings, and reduced urban water demand accounted for a mere 3% of savings, according to Jaeger and colleagues.

The reservoirs, in other words, are where we need to focus our regulatory attention. But, could it be done? How malleable are those rule curves? And what does that mean for flooding?

As Jaeger and colleagues point out, the rule curves at the reservoirs are far more flexible than many might believe. The operators—and here one can clearly see the insight of law professor Amos—are actually allowed a good deal of legal discretion in how they assess the risk of flooding vs. the benefit of storage.

However, changing the rule curves, as the researchers point out, is tricky business and entails a reassessment of the risk involved in determining how to find that Goldilocks level of enough water to meet summer needs, especially ESA requirements, but not so much that managers would find themselves unable to avert a flood if, for instance, a large rainfall might require them to do so.

So, what can we learn from this study?

As the researchers point out at the end of their paper, too often policy decisions are made without considering the complexity of the world they are intended to interact with. Jaeger and colleagues not so subtly hint that this needs to change. And that change, in our opinion, begins with studies like this one and projections like WW2100.

Citation: Jaeger, William K., Adell Amos, David R. Conklin, Christian Langpap, Kathleen Moore, and Andrew J. Plantinga. “Scope and limitations of drought management within complex human–natural systems.” Nature Sustainability 2, no. 8 (2019): 710-717. https://doi.org/10.1038/s41893-019-0326-y.

Featured Image: “Willamette Valley View,” November 13, 2008. Photo Credit: Don Hankins, some rights reserved.)

Nathan Gilles is the managing editor of The Climate Circulator, and oversees CIRC’s social media accounts and website. When he’s not writing for CIRC, Nathan works as a freelance science writer. Other posts by this Author.