What Beaver Dams Do to City Streams: Lessons From Oregon's Tualatin River Basin

Interest in beaver-assisted restoration has been growing in the Tualatin River Basin of northwestern Oregon. To see what that would really mean, the U.S. Geological Survey (USGS) and Clean Water Services gathered data in 2016–17 and ran a set of studies. Their overall finding: beaver dams help in some ways and at some seasons, cause trouble in others, and what they do depends on the particular stream and the particular dams.

An American beaver. Photo: Erin Leahy, U.S. Geological Survey.

Where the beavers are

Counting what's there. Combining USGS field surveys in 2016 of dams, lodges and chew marks (signs of beavers feeding) with other organizations' observations from 2011 to 2019, the team inventoried more than 650 beaver dams and 100 sites with lodges or chew marks across the basin. The inventory isn't complete, but it shows where American beavers (Castor canadensis) were active in 2011–19.

Beaver dams and beaver activity in the Tualatin River Basin, with the two study reaches (green diamonds) and the Urban Growth Boundary (grey). Map: U.S. Geological Survey.

Estimating what could be. A modified Beaver Restoration Assessment Tool (BRAT), adjusted for the basin's terrain, estimated how many dams the streams could hold as of 2017:

Planting streamside vegetation has other benefits too, such as shading small streams and cooling them.

Estimated potential beaver dams per kilometer of stream, inside and outside the Urban Growth Boundary, under 2017 vegetation and with more streamside vegetation. Chart: U.S. Geological Survey.

Up close: two urban creeks

The detailed work focused on two reaches: Fanno Creek at Greenway Park and Bronson Creek between Kaiser and Saltzman Roads.

Water and floods

Hydraulic models compared the reaches with and without their dams. Beaver dams and ponds let the reaches hold more water temporarily during storms, but they did not substantially change how fast, or how much, water flows through. They also created a wider range of water depths and speeds than a dam-free reach. Stream slope, side channels and the ponds themselves shaped how much water was stored, and for how long.

Modeled water depths and velocities in the Fanno Creek reach at low and high streamflow, with and without dams: the dams add shallow and deep habitat and slow-velocity habitat. Charts: U.S. Geological Survey.

Sediment

Over 12 months in 2016–17, the dams and ponds trapped:

The gap reflected differences in the dams and the reaches, in how well each stream connected to its floodplain, and in streamside vegetation.

One large Fanno Creek pond — the "south pond," flooding about 8,000 m² — trapped more than 250 metric tons of sediment in 4 years. For comparison, Fanno Creek near where it joins the Tualatin River carries off an average of about 7,200 metric tons of suspended sediment over 4 years, so this one pond held back roughly 1/30 — about 3% — of it. A chain of ponds along a stream network could trap a great deal of the sediment moving through urban streams.

Summer temperature

In both reaches, the ponds warmed the water in summer, but differently:

How much a dam warms a stream depends on the pond's surface area and depth, on shade, and on how the stream exchanges water with groundwater.

How summer water temperature changed along the Fanno Creek and Bronson Creek reaches: warming at the ponds, cooling where channels were narrow, deep and shaded. Diagram: U.S. Geological Survey.

Oxygen across a single pond

Measuring temperature and dissolved oxygen — the oxygen available to fish and other aquatic life — throughout the Fanno Creek reach on four summer afternoons in 2016 and 2017 showed how much conditions vary within one pond:

• Shallow, sunny parts of the south pond: warm and highly productive; algae photosynthesizing pushed oxygen to very high, supersaturated levels.
• Narrow, deeper, shaded channels: cooler, with less oxygen.
• The deep, shaded main and old channels: little photosynthesis, while leaves and wood decaying in the sediment used up oxygen — low oxygen.

Water temperature and dissolved oxygen through the Fanno Creek pond between 4:00 and 7:30 p.m. on August 11, 2016. Map: U.S. Geological Survey.

Streams built differently from Fanno Creek will behave differently again.

Benefits and trade-offs

The ponds also gave native turtles, amphibians and wading birds habitat in an urban landscape, and they can offset some of what urbanization does to streams — lost streamside vegetation, channelization, altered flow and sediment movement. But the benefits come with costs:

Photo: Erin Leahy, U.S. Geological Survey.

What it means for restoration

• Streamside vegetation is the main physical limit on beaver dams along these urban streams. Combining the dam model with maps of land use and infrastructure can pick out where beavers are welcome: reaches beside private land, roads, trails or culverts may suit beaver-assisted restoration less than reaches surrounded by public land with little infrastructure.
• General statements about beaver dams are true but not enough. Effects vary widely in kind and size, from season to season and place to place, even along one reach. Realistic expectations come from looking at the specific stream, dams and ponds and how they interact over time.
• Weigh benefits against challenges — such as flooding of property or infrastructure and warmer summer water — and build both into restoration plans, to limit conflict between people and beavers, explain the trade-offs and get the most out of a project.

The full studies

Four USGS Scientific Investigations Reports hold the details: Chapter A (White and others, 2025a) on where the dams are and how many the streams could support; Chapter B (White and others, 2025b) on storm flows and hydraulics; Chapter C (Doyle and others, 2025) on suspended sediment; and Chapter D (Smith and others, 2025) on water-quality patterns along the streams, across ponds and through the seasons.

Sources

Based on Jones K, Smith C, White J, Rounds S, Doyle M, Leahy E, "Beaver dams and their effects on urban streams in the Tualatin River Basin, northwestern Oregon," U.S. Geological Survey Fact Sheet 2025-3022; a work of the United States government in the public domain. The fact sheet describes the pond's 250 metric tons as a share of Fanno Creek's "annual" load while giving the 7,200-ton figure as a 4-year average; both figures are given here as 4-year amounts, without the word "annual". The photographs, maps and charts are reproduced from the fact sheet.