[1] Based on known relationships of slope, discharge, valley confinement, sediment supply, and sediment caliber in controlling channel patterns, we developed multivariate models to predict natural channel patterns across the 674,500 km² Columbia River basin, USA. We used readily available geospatial data sets to calculate reach slopes, 2 year flood discharge, and valley confinement, as well as to develop hypothesized landscape-level surrogates for sediment load and caliber (relative slope, percent of drainage area in alpine terrain, and percent of drainage area in erosive fine-grained lithologies). Using a support vector machine (SVM) classifier, we found that the four channel patterns were best distinguished by a model including all variables except valley confinement (82% overall accuracy). We then used that model to predict channel pattern for the entire basin and found that the spatial distribution of straight, meandering, anabranching, and braided patterns were consistent with regional topography and geology. A simple slope-discharge model distinguished meandering channels from all other channel patterns, but did not clearly distinguish braided from straight channels (68% overall accuracy). Addition of one or more of the hypothesized sediment supply surrogates improved prediction accuracy by 4–14% over slope and discharge alone. Braided and straight channels were most clearly distinguished on an axis of relative slope, whereas braided and anabranching channels were most clearly distinguished by adding percent alpine area to the model.

A pressing question for managing recovery of depressed or declining species is: Can habitat restoration increase resilience to climate change? We addressed this question for salmon populations with varying life histories, where resilience is defined as maintaining or increasing population size despite climate change effects. Previous studies indicate that several interrelated mechanisms may influence salmon resilience to climate change, including improving either habitat capacity or productivity, and ameliorating climate change effects on flood flow, low flow, or stream temperature. Using the Habitat Assessment and Restoration Planning (HARP) model, we first examined the relative importance of each mechanism for increasing salmon population resilience by comparing projected salmon spawner abundance for seven individual restoration action types under current and projected mid‐ and late‐century climates. We found that restoring habitats with the greatest restoration potential most increased resilience for all species, but the most beneficial restoration actions varied among species. Increasing habitat capacity and productivity both contributed to resilience, and ameliorating climate change effects was important in a few subbasins where the restoration opportunity was widespread. Cool‐water climate refuges contributed to resilience of some subpopulations by reducing late‐century declines in spawner abundance even without restoration. We also modeled more complex habitat restoration strategies comprised of several restoration action types at varying restoration intensities and found that the restoration action types and level of restoration effort needed to increase resilience varied among species. Less vulnerable species such as coho salmon responded well to four restoration actions (floodplain reconnection, wood augmentation, increased shade, and increased beaver ponds) applied at low restoration intensity and over a large area. More vulnerable species such as spring Chinook responded to fewer action types (floodplain reconnection, wood augmentation, and increased shade), but at much higher intensity and over a much smaller area. The analysis also identified important locations for each restoration action type for each species, which helps focus habitat restoration effort on areas that are likely to provide the largest increases in resilience.

Two approaches to ecological restoration planning, limiting-factors analysis and process-based restoration, are employed in efforts to recover endangered salmonid species throughout the Pacific Northwest of North America. Limiting-factors analysis seeks to identify physical limitations to fish production that may be addressed by habitat restoration; it is known as the “Field of Dreams” hypothesis (i.e., if you build it, they will come). Process-based restoration, in contrast, assumes that protection and/or restoration of watershed-scale processes will best achieve self-sustaining habitat features that support salmon populations. Two case studies from the Columbia River basin (northwestern USA) display current efforts to integrate these two restoration approaches to improve salmonid populations. Although these examples both identify site-specific habitat features to construct, they also recognize the importance of supporting key watershed processes to achieve restoration goals. The challenge in advancing the practice of restoration planning is not in simply acknowledging the conceptual benefits of process-based restoration while maintaining a traditional focus on enumerating site-specific conditions and identifying habitat-construction projects, but rather in following process-based guidance during recovery planning and, ultimately, through implementation of on-the-ground actions. We encourage a realignment of the restoration community to truly embrace a process-based, multi-scalar view of the riverine landscape.

Identifying necessary stream and watershed restoration actions requires quantifying natural potential habitat conditions to diagnose habitat change and evaluate restoration potential. We used three general methods of quantifying natural potential: historical maps and survey notes, contemporary reference sites, and models. Historical information was available only for the floodplain habitat analysis. We used contemporary reference sites to estimate natural potential habitat conditions for wood abundance, riparian shade, main channel length, and side channel length. For fine sediment, temperature, and beaver ponds we relied on models. We estimated a 90% loss of potential beaver pond area, 91% loss of side-channel length, and 92% loss or degradation of floodplain marshes and ponds. Spawning habitat area change due to wood loss ranged from -23% to -68% across subbasins. Other changes in habitat quantity or quality were smaller—either in magnitude or spatial extent—including rearing habitat areas, stream temperature, and accessible stream length. Historical floodplain habitat mapping provided the highest spatial resolution and certainty in locations and amounts of floodplain habitat lost or degraded, whereas use of the contemporary reference information provided less site specificity for wood abundance and side-channel length change. The models for fine sediment levels and beaver pond areas have the lowest reach-specific certainty, whereas the model of temperature change has higher certainty because it is based on a detailed riparian inventory. Despite uncertainties at the reach level, confidence in subbasin-level estimates of habitat change is moderate to high because accuracy increases as data are aggregated over multiple reaches. Our results show that the largest habitat losses were floodplain and beaver pond habitats, but use of these habitat change results in salmon life-cycle models can illustrate how the potential benefits of alternative habitat restoration actions varies among species with differing habitat preferences.