Specific Problem and Technical Background

Our goal is to restore healthy and harvestable salmonid populations through rehabilitation of stream habitat and restoration of ecological function in the riparian corridors of streams upstream of Grand Coulee Dam within the boundaries of the Colville Confederated Tribes reservation. Land use practices including residential and agricultural development, road building and logging have degraded habitat conditions and blocked upstream passage throughout many of the watersheds where historic tribal fishing once occurred. In order to identify and prioritize the appropriate fixes, we first need to inventory and understand the magnitude of the causes of that degradation. With that assessment in hand, we can form a plan to treat the root cause of the habitat degradation, rather than reacting to the symptoms which are observed at any one site.

The following excerpt was taken from “A Biological Strategy to Protect and Restore Salmonid Habitat in the Upper Columbia Region” (UCRTT 2008). It is our assumption that restoring ecological processes and allowing the streams to create their own habitat will be the most effective and efficient approach in the long term.

“Understanding ecological processes:  Many restoration projects fail because natural processes operating at different spatial and temporal scales and how human activities affect these processes are not well understood or considered. Implementation of successful restoration projects requires an understanding of these natural processes and the factors that control them (Frissell and Nawa 1992; Roni et al. 2002). Because these factors and processes operate at different spatial and temporal scales, restoration ecologists need to view the river holistically as a continuous “riverscape” (Fausch et al. 2002). The idea is that ecosystem processes operating at different scales form a nested, interdependent system where one level influences other levels. Thus, an understanding of one level is greatly informed by those levels above and below it. Furthermore, many processes that create habitat operate on time scales of decades or longer (e.g., channel migration and the formation of off-channel habitat) (Leopold et al. 1992). Interrupting natural ecosystem processes can result in the loss of fish habitat over the long term.

In simple terms, one can view the riverscape at three interconnected spatial scales: the geographic scale, the watershed scale, and the habitat/reach scale (Naiman et al. 1992; Montgomery and Buffington 1998). At the geographic scale, factors such as geology, soils, vegetation, and climate serve as ultimate controls (Leopold et al. 1992; Montgomery and Bolton 2003). These factors operate over large areas, are stable over long time periods, and act to shape the overall character and attainable conditions within a watershed or basin. Factors at the watershed scale are a function of geographic-scale factors and refer to more local conditions of geology, landform, and biotic processes that operate over smaller areas and shorter time periods. These factors include processes such as stream flows, temperature, sediment input, and channel migration. Factors operating at both the geographic and watershed scales help to define flow (water and sediment) characteristics, which in turn help shape habitat/reach-scale characteristics within broadly predictable ranges. Habitat/reach-scale factors include pool-riffle ratios, channel size, riparian vegetation, substrate composition, large woody debris, and bank stability. This is the scale at which fish species exploit resources and reproduce. This is also the scale at which most restoration occurs (Fausch et al. 2002).

Human activities that disrupt natural watersheds tend to act on processes that form suitable habitat conditions at the habitat/reach scale (Opperman et al. 2005) (Figure 3). For example, human activities can alter connectivity and the delivery of woody debris, water, sediment, and nutrients to a stream (Gregory et al. 2003; Stockner 2003; Opperman et al. 2005). Interruption of these processes reduces habitat quality and quantity at the habitat/reach scale by decreasing spawning and rearing space, food, and migration corridors. Likewise, restoration actions can focus on watershed processes or on habitats themselves (Figure 3). For example, some restoration techniques, such as re-vegetation, road removal, and establishing normative stream flows focus on restoring natural processes at the watershed scale. These techniques affect sediment supply, delivery of organic material, and channel migration. In contrast, other techniques focus on manipulating or enhancing habitat directly. Examples include wood and boulder placement, nutrient enrichment, and creating new habitat (Gregory et al. 2003; Stockner 2003; Morley et al. 2005). Unless well planned, with an in-depth understanding of ultimate controls and processes across different spatial and temporal scales, most habitat-enhancement techniques tend to be relatively short lived if the underlying process that has been disrupted is not corrected (Fausch et al. 2002).

In summary, successful restoration requires a holistic approach that considers processes operating at different spatial and temporal scales (Beechie and Bolton 1999; Beechie et al. 2010). A watershed or ecosystem assessment of current and historical conditions and disrupted processes is necessary to identify restoration opportunities that are consistent with reestablishing the natural processes and functions that create habitat (Roni et al. 2002). It is also essential to determine what restoration actions to implement first and how to prioritize actions (Roni et al. 2002). In general, restoration of watershed processes should precede or be conducted in conjunction with habitat enhancement. This is not to say that habitat enhancement techniques are inappropriate, but rather to emphasize the importance of coupling enhancement efforts with restoration of watershed processes. Clearly, in some locations (e.g., heavily urbanized areas) restoration of watershed processes may not be feasible. Habitat-enhancement techniques may be the only solution in these areas. In other areas, habitat enhancement techniques fall within the context of watershed processes and therefore are appropriate restoration measures.”

 Figure 1. Simple model showing linkages between landscape controls and watershed processes, and how land use and restoration or enhancement can influence habitat and biota (modified from Roni 2005) and taken from (UCRTT 2008).

Implementation Strategy and Scientific Justification

It is our intention to implement a restoration strategy that will restore and be consistent with the ecological processes described in the UCRTT Biological Strategy and the process-based principles outlined in Beechie et al. (2010). To accomplish this, our implementation strategy will follow the basic hierarchical strategy outlined in Roni et al. (2002). This approach begins with an assessment and inventory of the habitat conditions and degradations. Where habitat is degraded, we will seek first to reconnect isolated habitats then restore processes such as riparian condition and floodplain function before implementing actions that build temporary habitat. We will follow the recommended sequencing of actions from Roni et al. (2008) which is:

 1)  Protect high quality habitats

1. Functional habitats
2. Natural areas
3. Refuge areas

2)  Water quality and quantity

1. Improve water quality
2. Provide adequate flow

3)  Restore watershed processes

1. Habitat connectivity
2. Sediment and hydrology
3. Riparian and floodplains

 4)  Improve instream habitat

1. Instream structures
2. Nutrient enhancement

Without the inventory and assessment in hand it is impossible to describe what level of effort will be required under each of the techniques listed above. If known high quality fish habitats are at eminent risk of development or degradation we will employ habitat protection techniques such as land acquisition or conservation easements. Based on the work of previous projects we know that there are a considerable number of culvert barriers and an extensive history of grazing in the riparian zone. Therefore our project will begin implementation of actions under an opportunistic paradigm while the assessment is occurring. With the assessment in hand (year 3) we will re-focus to a more strategic approach, working down through the hierarchy of sequenced actions. Obviously, if there are apparent lethal flow or water quality problems those areas will be screened out of the list of potential projects for early implementation. Eventually, in unconfined areas where the channel migration zone is unimpeded and the stream has access to its floodplain but riparian vegetation is still recovering we may consider restoring instream habitat for short-term benefits through the addition of large woody debris (LWD). Additionally, in areas that have been heavily impacted by grazing or if rip rap is being removed, some LWD additions may be necessary as part of a treatment that accelerates the recovery of the damaged riparian vegetation and streambank.

Effectiveness and status and trend monitoring.—Although, we recognize the logic behind a comprehensive effectiveness and status and trend monitoring program, we do not believe that it is necessary or efficient to implement such a program in every watershed where habitat actions are occurring. We will primarily rely on other M&E efforts that BPA and their partners are implementing in other watersheds to determine the effectiveness of habitat restoration actions. For example, the Columbia Habitat Monitoring Program (CHaMP) is being implemented in 26 watersheds throughout the Interior Columbia Basin (Bouwes et al. 2011) and the Okanogan Basin Monitoring and Evaluation Program (OBMEP) is being implemented in the Okanogan basin to evaluate habitat status and trends (OBMEP 2011). Additionally, there are Intensively Monitored Watersheds in the Entiat, John Day, and Lemhi Rivers as well as intensive reach scale monitoring occurring in the Methow River. Presumably, the information gained from implementing these intense monitoring efforts will be applicable to other areas such as the San Poil River and other small streams of our study area. If that turns out not to be the case and managers want watershed specific habitat status and trend or effectiveness monitoring in our study area then we will implement those studies at a later date. For now, our approach is to collect initial baseline data in a manner consistent with CHaMP for the first 3 years of implementation for select attributes (water quality and quantity and macroinvertebrates, and fish density). Simultaneously we will conduct a systematic habitat inventory and assessment using the ODFW protocol to inform our project identification and selection and provide the basic habitat data needed to understand conditions affecting fish and to populate the EDT model. After focusing on implementation for the remainder of this review cycle (through 2017) we can reassess if additional habitat status and trend evaluation is warranted in our study area.

History, background, and review of past projects

The area that this proposal addresses contains the rivers and streams that occur on the Colville Confederated Tribes reservation and are tributaries to Lake Roosevelt. Among these are the San Poil River and its tributary streams, as well as smaller drainages on the eastside of the Reservation within the Upper Columbia Subbasin. The following description of the San Poil River drainage is from the San Poil Subbasin Plan (GEI Consultants Inc. 2004).

The San Poil River originates in the Okanogan Highlands east of the Okanogan River and drains in a southerly direction for 27 miles through parts of the Colville and Okanogan National Forests in Ferry and Okanogan counties. The river then enters the Colville Indian Reservation and flows approximately 32 miles south before it enters the impounded Columbia River in the San Poil arm of Lake Roosevelt at river mile 615.5.

The San Poil drainage forms the Water Resource Inventory Area (WRIA 52) as definedby the WDOE. The Subbasin encompasses approximately 981 square miles of Ferry and Okanogan counties (WDOE GIS data), which includes about 500 square miles of Tribal land on the Colville Indian Reservation. Elevations within the Subbasin range from 7,135feet above sea level at Copper Butte to 1,290 feet for Lake Roosevelt at full pool. Major tributaries to the San Poil River include Bridge, Gold, Granite, Iron, Louie, Lost, Manilla, Ninemile, North Nanamkin, O’Brien, Scatter, Thirteenmile, Seventeenmile, South Nanamkin, Thirtymile, Twentyfive mile, Twentythree mile creeks and the North, South, and West Forks of the San Poil River.

There are eight primary tributary streams that flow into Lakes Roosevelt on the eastside of the Reservation: Threemile, Sixmile, Ninemile, Wilmont, Nez Perce, Stranger, Hall, and Barnaby creeks. All are relatively small (4^th Order or smaller). Little is known about the habitat condition and fish distribution in these streams; although it is safe to surmise that there has been habitat degradation due to land use practices.

The Columbia River redband trout Oncorhynchus mykiss gairdneri are a subspecies of rainbow trout native to the Columbia River drainage east of the Cascade Mountains as far as barrier falls on the Snake, Spokane, Pend Oreille, and Kootenai Rivers (Allendorf et al. 1980; Behnke 1992). There is considerable variability in the life history of Columbia River redband trout (hereafter referred to as redband trout) including both anadromous (steelhead) and potamodromous forms (Behnke 1992). The San Poil River historically supported large runs of steelhead (Bryant and Parkhurst 1950; Fulton 1970), although the anadromous life history was eliminated with the construction of Grand Coulee Dam in 1939. The San Poil River steelhead population was part of the Upper Columbia ESU (ICBTRT 2003; McClure et al. 2005). It is likely that small spawning aggregates of steelhead and potamodromous redband trout also used various suitable habitats in many or all of the small tributaries that are on the Colville Confederated Tribes reservation. Studies have shown that there are fluvial, fluvial-adfluvial, and lacustrine-adfluvial populations of O. mykiss in the San Poil River drainage and many of the other streams on the Colville Confederated Tribes reservation that flow directly in to Lake Roosevelt. Some of these populations consist of genetically pure native redband trout, while others show some evidence of hybridization with coastal origin rainbow trout O. mykiss irideus that have been stocked throughout the upper Columbia River drainage (Powell and Faler 2002; Small and Dean 2007; Small and VonBargen 2009; Small and Bell 2011). Undoubtedly the population dynamics and life history pathways were substantially altered when steelhead were blocked from these tributaries due to the construction of Grand Coulee Dam.

Qualitative Habitat Analysis (QHA) was used to assess the limiting factors for fluvial, fluvial-adfluvial, and lacustrine-adfluvial redband trout in the rivers and streams of the San Poil and Upper Columbia subbasins during subbasin planning in 2004. Within the San Poil Subbasin obstructions was identified as the greatest problem for fluvial and fluvial-adfluvial redband trout, followed by riparian condition and habitat diversity. Low flows was most often indicated for lacustrine-adfluvial redband trout, followed by obstructions and high flow. Within the Upper Columbia Subbasin habitat diversity was identified as the greatest problem for fluvial and fluvial-adfluvial redband trout, followed by riparian condition and obstructions. Habitat diversity was most often indicated for lacustrine-adfluvial redband trout, followed by obstructions and fine sediment.

The Lake Roosevelt Rainbow Trout Habitat/Passage Improvement Project (LRHIP; BPA Project No. 1990-018-00) was initiated in 1990 to increase survival of migratory rainbow trout and ultimately improve Tribal subsistence and recreational harvest opportunity (LeCaire and Peone 1991). The project was implemented in three phases. Phase I (1990-1991) consisted of baseline habitat and fish population assessments that were conducted on 21 streams in the upper Columbia River drainage in Washington, 16 of which were Colville Confederated Tribes reservation. Following the baseline assessments, four streams (all in the San Poil River drainage) were prioritized for habitat restoration action. Phase II (1992-1994) of the project consisted of implementing habitat improvements, such as riparian planting, riparian fencing, instream structures, and barrier removals (Jones 2000). Phase III (1994-2000) entailed monitoring and evaluation of the habitat improvements (Jones 1999, 2000; Sears 2001). Monitoring of fish (juvenile out-migrant trapping, spawner escapement [weir], and backpack electrofishing) and habitat was conducted on these four streams. Then in 2001, it was concluded that the restoration activities in the initial four streams were relatively successful and the project began to focus habitat assessments, fish population monitoring (juvenile trapping, adult escapement, and backpack electrofishing), and habitat restoration projects on other streams (Sears 2001). Additional components were also added, such as genetic studies of small populations of rainbows, redd surveys, barrier assessments, and temperature monitoring.

After 2000, the intent of the project was to repeat the three-phase process on a different set of streams (Sears 2003). Although, the approach was altered in that habitat assessment was conducted on only one stream each year rather than a suite of streams followed by the use of prioritization criteria to identify restoration and subsequent monitoring activities (Sears 2003, 2004, 2005, 2007, 2008, 2009). The prioritization process that was used for stream selection after 2000 is unclear. Habitat surveys were conducted on nine streams in the San Poil River drainage between 2001 and 2011; however, they did not follow a standard protocol so the results are difficult to evaluate and we do not believe it would be scientifically valid to try to use those results as a baseline for identifying restorations projects or long-term habitat status and trend monitoring. This project did allow personnel to get familiar with the watershed and identify some opportunities for implementing habitat restoration projects. However, under new leadership, we plan to follow the scientific approach to habitat monitoring and restoration outlined in the previous section of this proposal.

Habitat improvement projects implemented under the LRHIP between 2001 and 2011 included the replacement and removal of culverts, channel reconstruction, bank stabilization, restoring floodplain connectivity, riparian fencing, riparian planting, nutrient enhancement, and transfer of water rights (Sears 2003, 2004, 2005, 2007, 2008, 2009; CCT, unpublished data). The bulk of the habitat improvements were completed in the San Poil River drainage. Evaluation of restoration project efficacy generally consisted of monitoring adult escapement prior to and post completion of the restoration action. There was no significant increase in adult abundance (counts) pre- and post- habitat restoration in the San Poil River drainage streams examined, although inference was limited due to small sample sizes (Buchanan 2009). Additional monitoring was needed to detect a difference between pre- and post- restoration counts. Additional monitoring was completed on some streams, but the data has not been analyzed.

 As previously mentioned, the LRHIP project also conducted other monitoring activities; however the relationships of these activities to habitat restoration activities and status and trend monitoring are not clear. Fish monitoring was completed (adult escapement) on various other streams that had not had habitat improvements, presumably to assess population productivity and examine life histories. Small-scale genetic studies were conducted to determine the level of hybridization between redband trout and introduced coastal rainbow trout (Powell and Faler 2002; Small and Dean 2007; Small and VonBargen 2009; Small and Bell 2011). The genetic studies included samples from fish captured in numerous reservation streams, although samples sizes from individual populations were small. As previously mentioned, some of these populations appear to consist of genetically pure native redband trout, while others show some evidence of hybridization with coastal origin rainbow trout. Redd surveys were completed; however, the data are not available and the reason behind the surveys has not been clearly articulated.

Salmon carcass analogs were distributed in ten streams in the San Poil River drainage in 2008 and 2009 to enhance nutrient levels (Sears 2009; CCT, unpublished data). Carcass analogs were also distributed in three of the original ten streams in 2010. Nutrient enhancement was evaluated by comparing macroinvertebrate assemblages and abundance both upstream and downstream of treatment reaches in three streams in 2009 and 2010. There was not a significant increase in macroinvertebrate abundance following the nutrient treatments and the low number of samples resulted in poor statistical power (Richards 2011).

There are numerous man-made and natural barriers to fish passage on CCT reservation streams, but a general lack of information limits our ability to identify and prioritize passage improvement projects. An objective of the LRHIP was to conduct an inventory and assessment of all man-made and natural fish passage barriers (Sears 2003, 2004, 2005, 2007, 2008, 2009); however, the methods consisted of professional judgment and the information that was recorded is currently not available. Tribal road inventories identified barriers to fish passage at road crossings, although it is unclear what criteria were used during the assessment and not all drainages have been completed (DCA 2004, 2005, 2006, 2008, 2009, 2011). Although, the existing road crossing information will aid with filtering those crossing locations that will not require re-visiting for a fish passage assessment (i.e. those high in the watershed and dry at low flows). Once we have filtered the sites, we will systematically inventory and assess the remaining potential barriers to passage for all life stages of fluvial-adfluvial and lacustrine-adfluvial rainbow trout using a standard protocol.

Most recently, ICF International was subcontracted to conduct a watershed assessment for the east side tributaries (still in draft) (Figure 2) using a combination of remote sensing (GIS) data and some on the ground information (tabular) for water quality and discharge that was available from various sources but was not part of a systematic or probabilistic sampling design and did not follow a standard protocol (Table 1).

Table 1. Summary of watershed analysis output parameters and the watersheds evaluated for the east side tributaries of the Colville reservation. 

The land cover analysis determined that nearly 1/3 of the study area were either pasture (16%) or clearcut/conifer regeneration (15%) (Figure 3). This supports the assumption that land management practices are likely having meaningful effects on the stream corridors and instream conditions. Likewise, the temperature regime in Hall Creek often exceeded the preferred temperature range for salmonids between May and September (Figure 4). However, Barnaby Creek rarely exceeded 20 degrees Celsius, indicating a much better thermal regime for rearing salmonids (Figure 5).

Initial recommendations from this assessment were that Hall and Barnaby creeks were the highest priority followed by Wilmont and Ninemile creeks. Unfortunately, Wilmont and Ninemile creeks are known to have natural barriers very near their confluence with Lake Roosevelt so we are excluding them from our consideration for restoration because we are prioritizing the lacustruine-adfluvial life history due to their size and Tribal interest in these large fish. Threemile and Sixmile creeks had limitations regarding watershed size, low flow, and high gradient so they were also excluded from our consideration. Therefore we will be focusing our eastside tributary efforts in Barnaby, Hall, Stranger, and Nez Perce creeks.   

A similar analysis is currently underway for the San Poil River drainage.

Figure 2. The primary Upper Columbia Subbasin stream drainages on the eastside of the Colville Confederated Tribes reservation.

Figure 3. Landcover types within the Upper Columbia Subbasin stream drainages on the eastside of the Colville Confederated Tribes reservation.

Figure 4. Water temperatures in Hall Creek, an Upper Columbia Subbasin stream on the eastside of the Colville Confederated Tribes reservation.

Figure 5. Water temperatures in Barnaby Creek, an Upper Columbia Subbasin stream on the eastside of the Colville Confederated Tribes reservation.

Management questions

All of the subsequent management questions refer to our study area which is the tributary creeks of Lake Roosevelt within the boundaries of the Colville Confederated Tribal Reservation including the San Poil River, Barnaby Creek, Hall Creek, Stranger Creek, and Nez Perce Creek.

1)  What are the current habitat conditions?

2)  Where are the natural barriers to upstream passage of adfluvial salmonids?

3)  What are the root causes of habitat degradation in the streams and riparian corridors?

• Sub question #3a. Where and how extensive are the passage barriers for fluvial-adfluvial and lacustrine-adfluvial salmonids?
• Sub question #3b. Where and how extensive are the causes of water quality perturbations?
• Sub question #3c. Where and how extensive are the instream flow limitations?
• Sub question #3d. Where and how extensive are the riparian vegetation and stream bank degradations?
• Sub question #3e. Where and how extensive are the disruptions of the channel migration zone, floodplain, and sidechannel connectivity?

4) What are the highest priority restoration actions to improve salmonid production?

5) What is the current distribution and density of fish?

Key project personnel

Jason McLellan—Project Manager. Master of Science degree in Biology (Fisheries emphasis) from Eastern Washington University. More than 13 years’ experience managing and implementing a diverse array of research and monitoring projects focused on resident fish conservation and management in the upper Columbia River drainage. Primarily focused efforts on native resident fish, such as white sturgeon, burbot, redband trout, and westslope cutthroat trout. Designed and implemented habitat and fish population assessments in small streams and rivers. Participated in watershed planning activities in the Spokane and Colville River drainages, as well as the FERC re-licensing process in the Spokane and Pend Oreille River drainages.

Josh Hall—Lake Roosevelt Habitat Improvement Project Biologist, Bachelor of Science degree in Natural Resource Sciences from Washington State University, five years of on the ground experience doing fish and habitat surveys in streams on the Colville Indian Reservation. Project lead for restoration project implementation from 2008 to present. Participates in multi-disciplinary natural resource project reviews as part of an internal review process for the Tribe. Participated on the San Poil Watershed Action Team for habitat project prioritization in the San Poil and Upper Columbia Sub-basins in conjunction with Conservation Districts, United States Forest Service, Washington Department of Fish and Wildlife, and others. Knowledge of PIT tag and radio tag implanting in juvenile salmonids and installation of necessary hardware (i.e. antennas, in-stream PIT tag arrays, etc.) for fish passage studies on the Snake River while working for NOAA fisheries 2005-2007.

Lucas Dailey—Lake Roosevelt Habitat Improvement Project, Fish Biologist 1. Bachelor of Science degree in Natural Resource Sciences from Washington State University, four years of diverse fish and wildlife experience ranging from spotted owl surveys to conducting ungulate habitat assessments. Participated in numerous biological studies in parts of Alberta, Idaho, Oregon, Washington, and Wyoming. Eighteen months of experience performing fish population and habitat assessments, including 12 months on the ground experience in Lake Roosevelt tributaries on the Colville Indian Reservation. Field supervisor for implementing restoration projects and studies pertaining to resident fish habitat on the reservation.

Casey Baldwin—Science and planning oversight and consultation. Master of Science degree in Fisheries from Utah State University, 9 years experience in the Upper Columbia Tributaries including 4 years as chairperson of the Regional Technical Team. Evaluated hundreds of fish habitat restoration project proposals for Salmon Recovery Funding Board, Mid-Columbia Tributary Committees, and Bonneville Power Administration Fish and Wildlife Program for the Columbia Cascade Province. Project lead on the EDT assessments for spring Chinook and steelhead in the Methow subbasin during the NPPC subbasin planning and the Wenatchee subbasin during the development of the salmon and steelhead recovery plan (UCSRB 2007). Participated on numerous habitat project development teams in coordination with Watershed Action Teams, Conservation Districts, the United States Bureau of Reclamation, and others. Participated on the 2006 and 2009 expert panel for the Upper Columbia watersheds to assist the Action Agencies with their process for assigning credit for the actions taken under the FCRPS BiOp.

Project Objective

The objective of the project is to restore healthy and harvestable salmonid populations through rehabilitation of stream habitat and restoration of ecological function in the riparian corridors of streams upstream of Grand Coulee Dam within the boundaries of the Colville Confederated Tribes reservation.