Summary

This study was a collaborative effort between the U.S. Geological Survey, The Nature Conservancy, the Eugene Water & Electric Board (EWEB), and the U.S. Army Corps of Engineers under the auspices of the Sustainable Rivers Project. In 2002 The Nature Conservancy and U.S. Army Corps of Engineers began the Sustainable Rivers Project for the purpose of modifying dam operations and implementing environmental flow requirements for various river systems around the country. This report provides a baseline assessment of McKenzie River basin hydrology, geomorphology, and biology and their linkages. Information from the report will assist McKenzie River basin stakeholders in the development of future environmental flow requirements at an upcoming workshop. Final decisions from the workshop will be published in a second report.

The McKenzie River is approximately 90 miles long with a drainage area of approximately 1,300 square miles. Streamflows originating in the upper basin are fed by highly productive springs that were formed as a result of the young High Cascade volcanic geology. For this study the river was divided into 12 study reaches, each having unique streamflow and sediment input conditions. The McKenzie River is regulated by two flood control projects (Cougar and Blue River), an upper basin hydropower complex (Carmen–Smith–Trail Bridge), and two hydropower canals (Leaburg and Walterville). The upper basin EWEB dams (Carmen, Smith, Trail Bridge) have slightly affected hydrology by decreasing annual 1-day maximum streamflows. However, the Cougar and Blue River dams have significantly reduced the magnitude and frequency of floods (on average, frequency of small floods reduced by 46 and 51 percent, respectively) while increasing the annual 7-day minimum streamflows. In the lower reaches, the McKenzie River supplies streamflow to the Leaburg and Walterville power canals. Up to 2,500 ft³/s of the streamflow can be diverted into the canals so long as a minimum of 1,000 ft³/s of streamflow remains in the McKenzie River.

Sufficient streamflows and cool stream temperatures are essential habitat requirements for aquatic species in the McKenzie River basin. In 2006 the Oregon Department of Environmental Quality listed a total of 112.4, 6.3, and 55.7 miles of McKenzie River basin mainstem and tributary stream reaches as thermally impaired for salmonid rearing, salmonid spawning, and bull trout habitat, respectively. Stream temperatures are affected by a variety of anthropogenic factors, such as the removal of riparian vegetation, urban and agricultural land use, geomorphic disturbance, and dam streamflow releases. Although the effect of upper basin dams on stream temperatures has been minimal, streamflow releases from the Cougar and Blue River dams have been unnaturally cooler in the summer and warmer in the fall. These changes to the thermal regime can be observed downstream on the McKenzie River at the streamflow gage near Vida, Oregon.

McKenzie River geomorphology was evaluated using reach characterization, historical channel mapping, and specific gage analysis methods. Although more detailed studies are needed to determine the precise effects of various magnitude flow events on physical habitat in different areas of the McKenzie basin, the findings from this study, in combination with earlier studies, provide a framework for assembling general relationships between streamflow and physical habitat.

In the upper (Reaches 1–2) and middle (Reaches 3–8) McKenzie River basin, most pools are formed at bedrock outcrops. In contrast these forced pools are less important in the lower McKenzie River basin (Reaches 9–12) than scour pools that typically resulted from channel-spanning blockages of wood and sediment. Pool complexity depends on multiple factors, including availability of large wood, overhanging banks, and sediment supply. Discharge events ranging in magnitude from bankfull streamflow to small floods likely could improve pool habitat throughout the middle and lower basins of the study area by deepening pools, flushing fine sediments, and potentially recruiting large wood that would enhance cover. Large floods could, in addition to improving pool habitat, potentially create new pools, especially in the lower McKenzie River basin where channel shifting and island growth can facilitate pool formation.

Offchannel habitat historically was most abundant along the lower, alluvial reaches of the McKenzie River basin, whereas the steeper, more confined reaches along the middle and upper McKenzie River basins presented few opportunities for secondary channel features. Therefore, discharge events that exceed bankfull streamflow are likely to have the greatest effect on secondary channel features in the lower McKenzie River basin by enhancing side channels, alcoves, and sloughs. Small floods may scour and maintain existing lower elevation features in active channel, but larger magnitude floods are needed to actually carve and create new secondary features, particularly along floodplain swales and older higher elevation surfaces in the active channel.

Depositional features, including gravel bars and spawning gravels, historically were most abundant along the lower McKenzie River basin. Presently, the formation of gravel bars is limited by multiple factors including (1) sediment supply, resulting from naturally low sediment yield rates from the upper basin, and exacerbated by trapping of bed-material by dams, (2) high velocity transport capacity capable of transporting available gravel and cobbles to lower reaches rather than creating depositional zones where bars can form, (3) limited recruitment by bank erosion because of reduced flows and bank stabilization, and (4) vegetation colonization and stabilization of formerly active bar surfaces, resulting from reduction in peak flows. Previous studies have indicated that decreases in spawning habitat on the McKenzie River partly may be a result of bed coarsening, which can be exacerbated by sediment trapping behind dams; however, further studies are needed to determine relations between bed substrate, spawning habitat, and sediment availability.

Streamflow plays an important role in determining the availability of gravel bars and other depositional features because it not only determines transport capacity, but also drives bank erosion and disrupts vegetation on gravel bars, hence, allowing for remobilization of gravels stored in bar and floodplain deposits. It can be hypothesized that bankfull streamflow events may reduce vegetation on low bar surfaces and may remobilize existing bare bars, but small to large floods are needed to substantially rework the heavily vegetated (but historically active) gravel bars. Furthermore, small to large flood events (2- to 10-year recurrence interval) will also have a greater capacity to trigger bank erosion, particularly along unrevetted areas of the lower McKenzie River basin alluvial reaches. This could release gravels and large wood that facilitate the creation of new bars and spawning habitat. In order to further refine these relationships, additional studies are needed to quantify sediment transport capacity and bank erosion under a range of flow conditions, and to also develop sediment budgets for different parts of the McKenzie River basin.

Alluvial reaches along the lower McKenzie River basin historically supported broad, forested floodplains, whereas the floodplains along the middle and upper McKenzie River basins are typically narrow and confined by steep valley walls. Reduction of peak streamflow has resulted in fewer overbank flood events that carve and maintain floodplain channels, which not only provide refuge during high streamflows, but also contribute to vegetative patch heterogeneity. Furthermore, flow reduction, in combination with bank stabilization and reductions in sediment supply, has decreased the frequency of “floodplain recycling events,” which occur when erosion of floodplain surfaces mobilizes sediment and large wood that are deposited elsewhere and eventually evolve into future floodplain surfaces. Therefore, the discharge events that likely are to have the greatest effect on floodplain formation are overbank events (such as small to large floods) when floodplain channels can be scoured and bank erosion can more substantially modify floodplain margins. However, at least one area along the lower McKenzie River basin seems to be experiencing incision, and if incision is widespread, it could negatively influence floodplain habitat by reducing connectivity between the channel and its adjacent floodplains. Additional studies are needed to evaluate the magnitude and extent of incision and to determine whether environmental flow releases would exacerbate bed lowering.

Nine exemplar aquatic and terrestrial species used in the study included spring Chinook salmon, bull trout, Pacific and brook lamprey, Oregon chub, red-legged frog, western pond turtle, white alder, and cottonwood. These species were combined into three groups based on the primary habitats they use: mainstem aquatic species (Chinook salmon, bull trout, and Pacific and brook lamprey), offchannel aquatic species (Oregon chub, red-legged frog, western pond turtle), and floodplain and riparian vegetation (white alder and black cottonwood).

Spring Chinook salmon, bull trout, and brook and Pacific lamprey are species native to the McKenzie River basin that require a diversity of instream and offchannel habitats. In addition, they are all coldwater species and spawn in similar flowing-water habitats. These four species use or require habitat features that are created and maintained primarily by the interaction of small to large floods with local landscape features. Habitat features of particular importance and common to these species include large, deep mainstem pools, secondary channels, and well sorted spawning gravels with interstitial streamflow. All four species also are sensitive to the seasonal timing of streamflows and water temperatures and thus are affected by alterations in streamflow and temperature associated with dam operations.

The Oregon chub, western pond turtle, and red-legged frog are native species that require a diversity of offchannel habitats, including mature side channels, sloughs, oxbow lakes, ponds, and wetlands. These kinds of habitats are created and maintained by the interaction of high streamflow events such as small to large floods with local landscape features. In addition, high streamflows in winter and spring are essential for the dispersal of seed and vegetative propagules of riparian and floodplain trees and shrubs. High streamflows also generate and maintain areas of bare soils necessary for germination, and their timing and ramping rates during reservoir releases are critical for the maintenance of proper water tables necessary for growth and survival of riparian vegetation.

The changes caused by the dams to natural streamflows affect all these and other species in complex ways; nevertheless, a few commonalities can be stated. The reduction in small to large floods in the McKenzie River basin, along with bank stabilization, likely are primary factors involved in reducing channel complexity, which has been evidenced by reductions in mainstem spawning sites, deep complex pools, and secondary channel features. Reductions in these key habitat types have been implicated as a key factor in the population declines of all nine exemplar aquatic and terrestrial species. The dams also have had direct effects by blocking access to habitat, changing the amount and timing of important transient habitats, and altering water temperature and its seasonal timing (important as migratory, growth, and life history cues), as well as affecting other factors important to the exemplar species outlined in this report.

First posted February 8, 2010