Skip to content
Ecological Intelligence

Ecological evidence.
Independent perspective.
Explore Wild Lands ↗

Ecological Intelligence / Subject

stream restoration

Peer-reviewed North America

After two decades of stream work, watershed-wide habitat signals were often missing

Evidence
Peer-reviewed
Geography
Western Washington, United States
System
Forest streams; salmonid habitat and restoration

WHAT THE EVIDENCE SAYS

From 2007–2024, investigators analyzed 4,421 habitat surveys from 269 sites across ten watersheds in three complexes, tracking wood, pools, gravel and channel geometry alongside varying restoration effort. Most expected whole-watershed directional shifts and treated-versus-reference differences did not emerge. Year-to-year and within-watershed variability were substantial.

WHY IT MATTERS / OUR READ

A single before-and-after snapshot cannot tell whether watershed restoration worked. Programs need repeated, spatially distributed measurements and a defined inference scale before investment can be assessed.

CAVEATS / UNCERTAINTY

Failure to detect a watershed-wide average habitat change is not proof that individual restoration projects had no local benefit or no salmon response. Some reference watersheds also experienced management; this paper assesses common habitat metrics, not every ecological objective.

SOURCE

Frontiers in Ecology and Evolution — primary long-term monitoring study
View original evidence ↗

Thousands of habitat surveys across ten Washington watersheds found interannual changes often larger than long-term trends or expected restoration contrasts.

Read the Wild Lands assessment →
Peer-reviewed North America

Beaver dam analogues helped bats and some amphibians—but pond formation was the threshold for breeding

Evidence
Peer-reviewed
Geography
Okanogan-Wenatchee National Forest, Washington Cascades, USA
System
Montane streams and riparian restoration

WHAT THE EVIDENCE SAYS

Researchers compared 25 beaver-dam-analogue sites with paired controls for amphibians, 18 BDA-control pairs for bats, and natural beaver sites as a reference. Amphibian breeding occupancy was 24% at BDA sites, zero at paired controls and 88% at natural beaver sites. Bat activity was 3.9 times higher at BDA sites than controls and 4.7 times higher at beaver sites than controls. Amphibian response depended strongly on whether the structures created ponded habitat, while bat activity tracked slower water.

WHY IT MATTERS / OUR READ

For managers, the treatment is the habitat function—not the number of structures installed. If amphibian breeding is an objective, pond formation and hydroperiod need to be monitored explicitly; if bats are an objective, slower water and expanded foraging habitat may matter. Natural beaver sites also remind managers that an analogue may reproduce selected functions without reproducing the full complexity of beaver engineering.

CAVEATS / UNCERTAINTY

The BDA sites were only one to two years old. Amphibian BDA-control comparisons were paired, but comparisons with natural beaver sites were not always paired; occupancy and acoustic activity are not population-growth or recruitment measures. Results depend on local hydrology, species, permits and infrastructure constraints and should not be universalized.

SOURCE

Restoration Ecology — primary study
View original evidence ↗

In Washington mountain streams, amphibians bred at 24% of BDA sites versus none of paired controls, while bat activity rose 3.9-fold; natural beaver sites still delivered more complete habitat for amphibians.

Read the Wild Lands assessment →
Peer-reviewed North America

Beaver-dam analogues helped coho persist through drought where neighboring populations collapsed

Evidence
Peer-reviewed
Geography
Scott River tributaries, northern California, USA
System
Drought-prone salmon streams and riparian wetlands

WHAT THE EVIDENCE SAYS

Longitudinal monitoring in Scott River tributaries evaluated beaver-dam analogues and associated off-channel habitat for coho salmon. Restored reaches retained more slow, cooler water during drought, supported thousands of additional juveniles, and in French Creek mean juvenile survival increased from roughly 8% before restoration to about 60% afterward. Adult returns also strengthened relative to neighboring unrestored systems.

WHY IT MATTERS / OUR READ

Process-based restoration can change what a drought means biologically. The structures did not remove regional drought; they changed local hydrology and habitat so fish had refuges when neighboring channels contracted.

CAVEATS / UNCERTAINTY

This is a watershed-scale before/after observational restoration study rather than a randomized experiment. Responses depend on geomorphology, groundwater, maintenance, fish access and drought severity, so effect sizes should not be generalized unchanged to other basins.

SOURCE

Frontiers in Ecology and Evolution · Original evidence
View original evidence ↗

NOAA Fisheries · Research summary
View research summary ↗

Longitudinal monitoring in California’s Klamath Basin linked restored slow-water habitat to cooler refuges, higher juvenile survival and stronger adult returns.

Read the Wild Lands assessment →

Keep exploring

Follow the systems. Find the connections.

WildlifeLandWaterOceansClimateCitiesOne HealthGovernanceMethods / Monitoring
Browse the archive →