Watershed Restoration Program

Ministry of Environment, Lands and Parks

Williams Lake, BC

K. Brydges, S. Luzzi, T. Roy and J. Horchik

of

March 1998

At the request of the British Columbia Ministry of Environment, Lands and Parks (MELP), Cariboo Region, Williams Lake, Carmanah Research Ltd. conducted a Level 1 Fish Habitat Assessment Procedure (FHAP) of Clisbako River and McFarland Creek. The FHAP is a process that identifies areas where fish habitat has been impacted by logging and recommends prescriptions for rehabilitation.

The procedures used during this survey followed those outlined in the Fish Habitat Assessment Procedures, Watershed Restoration Technical Circular No.8, April 1996 Revision (Johnston and Slaney, 1996). Site locations were initially chosen by selecting those reaches that were identified as high priority in the FHAP overview report of the Nazko River watershed (Carmanah Research Ltd., 1997).

The study area for the Level 1 FHAP was comprised of the Clisbako River mainstem and one third order tributary, McFarland Creek. Both are sub-basins in the Nazko River watershed. The study area is located on the Interior Plateau of central British Columbia. The Clisbako River is 109.5km long, and McFarland Creek is 16.4km long.

The main goal of the FHAP assessment was to identify sections of streams that have been impacted by logging, and to recommend prescriptions for Level 2 restoration work. During the Level 1 assessment in the Clisbako and McFarland watersheds, approximately 25.3km of stream were assessed. Descriptions of each reach are accompanied by verification of fish presence or absence, as well as remediation prescriptions.

The results of the Level 1 assessment indicate that the only area affected by logging was reach 1 of the Clisbako River. A logging road runs parallel to the channel and is likely contributing sediment to the stream. Free-range cattle roam the area and the banks have been damaged in reach 1 where they cross the stream. The disturbance indicators in this reach were a lack of pool habitat, and compacted bed material. Preliminary restoration prescriptions include complexing the habitat in the reach by adding LWD and boulder clusters, building a fence along the length of reach 1 to control livestock movement, and controlling sediment sources with cribwalls or by re-vegetating the banks.

The 1997 Level 1 Fish Habitat Assessment of the Clisbako River and McFarland Creek watersheds was directed by the Ministry of Environment, Lands and Parks, Cariboo Region, Williams Lake with funding provided by Forest Renewal BC. The primary contractor was Carmanah Research Ltd. of Victoria, B.C. with support and monitoring provided by Watershed Restoration Program Specialist Michael Parker (MELP). Invaluable field support was provided by M. Burrell and J. Touchie. Report preparation could not have been accomplished without the enthusiastic hard work put forth by K. Brydges, S. Luzzi, T. Roy and H. Liebe-McGinnis. Project guidance was provided by K. Brydges. Z. Gedalof produced the maps for this report.

Report quality was assured by S. Luzzi, D. Lafleur, T. Roy and P. Harder, R.P.Bio.

Executive Summary *

Acknowledgements *

Table of Contents *

List of Tables *

List of Figures *

List of Appendices *

1.0 Introduction *

2.0 Study Area *

3.0 Methods *

4.0 Results *

5.0 Discussion *

6.0 General Recommendations *

Literature Cited *

Table 1. Fish catch summary by sample site. *

Figure 1. Nazko Watershed *

Appendix A. Photographs of significant habitat features and reach-representative sites

In 1996 the Ministry of Environment, Lands and Parks (MELP), Cariboo Region, initiated a Watershed Restoration Program (WRP) for the Nazko River watershed, with funding provided by Forest Renewal British Columbia (FRBC). As an initial step in this process, a Fish Habitat Assessment Procedure (FHAP) was started. In 1996, Carmanah Research Ltd. was contracted by MELP to conduct an Overview FHAP of the Nazko watershed following the procedures outlined in the Fish Habitat Assessment Procedures, Technical Circular No.8 (Johnston and Slaney, 1996). From this report, recommendations were made to conduct more detailed surveys in specific areas of the watershed. One of these recommendations was to conduct a Level 1 assessment on reaches 1 through 4 of the Clisbako River and reaches 1 and 2 of McFarland Creek. It was also recommended that the falls at the top of reach 4 on Clisbako River be assessed to determine if they were a barrier to fish migration. Carmanah Research Ltd. was contracted in June 1997 by MELP to complete a Level 1 Fish Habitat Assessment of approximately 25.3km of stream within the Clisbako and McFarland watersheds.

The Nazko River watershed is located on the Interior Plateau of central British Columbia (Figure 1). The Clisbako River is a sub-basin of the Nazko River, located in the western portion of the Nazko River watershed (Figure 2). This river has a drainage basin area of 81,688ha and a stream length of 109.5km. Nineteen reaches were identified in the overview report (Carmanah Research Ltd., 1997). McFarland Creek, a third order tributary, joins the Clisbako River 9.2km upstream of the mouth. McFarland Creek has a stream length of 16.4km and was identified as having 19 reaches in the overview report (Carmanah Research Ltd., 1997). The overview report also identified a potential gradient barrier in reach 3 of McFarland Creek, where the slope reaches 21.9% (Carmanah Research Ltd., 1997).

Figure 1. Nazko Watershed

Figure 2. The Clisbako River Sub-basin is located in the Nazko Watershed (inset).

The Clisbako River watershed lies within the physiographic region of the Interior Plateau. The topography consists of gently rolling, undissected uplands interspersed with numerous small lakes and wetland areas. The bedrock geology is of volcanic origin, consisting of gently dipping, Miocene/Pliocene olivine basalt flows. The surficial materials are mainly unconsolidated glacial deposits. Fluvial and glaciofluvial deposits dominate the mainstem of the Clisbako River (Lord and Walmsley, 1988; Carmanah Research Ltd., 1997).

The climate of the study area is characteristically dry. The average annual temperature is approximately 2° C, with average summer highs in July-August of 14° C and January winter lows averaging -13° C. Average annual precipitation is approximately 440 mm, with an estimated 40% falling as snow. The majority of rainfall occurs in June, July and August.

The study area falls within the Sub-boreal Pine and Spruce (SBPS) and Montane Spruce (MS) biogeoclimatic zones. The SBPS zone is characterized by typically low productivity forests interspersed with wetland areas. Lodgepole pine is the primary tree species, with white spruce secondary. The MS zone is characterized by mixed spruce - lodgepole pine forests of relatively uniform stand age.

Water Survey Canada has maintained a gauging station (No. 08KF001) on the Nazko River upstream of Michelle Creek since 1965. The mean annual flow over the period of record is 4.61 m³s^-1. To estimate the mean annual flow for the Nazko River at the mouth, a conversion of the measured discharge by a ratio of the total watershed area (415 000 ha) over the gauged watershed area (324 000 ha) gives a discharge of 5.9 m³s^-1. Average maximum and minimum monthly discharges take place in May and January-February respectively. Summer and winter 7 day low flows calculated by Rood and Hamilton (1995) using records from 1981 to 1990 are 1.27 m³s^-1 and 0.96 m³s^-1 respectively. Low summer flows can affect fish habitat by increasing water temperatures, stranding juveniles and reducing rearing habitat. Low winter flows can dewater redds and endanger incubating eggs by increasing the probability of freezing (Rood and Hamilton, 1995).

Chinook spawning has been documented in reach 1 of the Clisbako River. Rainbow trout have also been identified within that reach (Appendix D). The overview report identified a falls at the upstream end of reach 4, observed during a helicopter reconnaissance flight of the watershed in October, 1996, which appeared to pose a barrier to fish migration. It is not known whether any resident fish are present upstream of the falls. The upper portions of the stream have been identified as poor fish habitat with limited fish production potential due to the low gradient and lack of gravels in the bed material (DFO Stream Information Summary, 1990; Rood and Hamilton, 1995; Carmanah Research Ltd., 1997). There are no fish inventory data for McFarland Creek; it is unknown whether any resident fish are located above the steep section in reach 3. Table 1 is a summary of the fish caught during the 1997 FHAP in each reach of the Clisbako River, along with two sections of McFarland Creek.

Table 1. Fish catch summary by reach for Clisbako River and McFarland Creek.

K - Presence noted visually

S - Suspected presence

The major land use activities within the study area are agriculture and forestry. Free-range cattle have been observed grazing along the lower reaches of the Clisbako River; fording cattle have contributed to bank erosion, and consequently sedimentation. This affected the stream in reach 1.

Extensive forest harvesting has occurred in the Clisbako and McFarland watersheds. This was primarily salvage logging that followed a pine beetle infestation from 1985 to 1989. The areas affected included reaches 1 through 3 on the Clisbako River, and portions of the McFarland Creek watershed (C. von Hahn, pers.comm., 1996). Large portions of the Clisbako sub-basin headwaters have been harvested between 1989 to 1990 (MOF, 1996). Leave strips were maintained along Clisbako River, but many first and second order tributaries have been logged to the stream bank. Past logging activities have also created a network of roads and crossings that were identified in the overview report as having an impact on the stream (Carmanah Research Ltd., 1997).

The fish habitat assessment of Clisbako River and McFarland Creek was conducted according to the procedures outlined in the Fish Habitat Assessment Procedures manual WRP Technical Circular No. 8 (Johnston and Slaney, 1996). Fish habitat assessments were completed for those reaches which had previously been identified as high priority in the FHAP Overview of the Nazko Watershed (Carmanah Research Ltd., 1997). These reaches were sampled using a sub-sampling fraction of 1/3 for each habitat unit. This fraction ensured a detailed analysis of about 33.3% for all pools, glides, riffles, and cascades. Some habitat unit types were sub-sampled at 1/2 and 1/1 when stream conditions warranted a higher fraction (i.e. very few pools). Habitat unit averages for the entire reach were calculated from values obtained in the detailed assessments. The detailed habitat surveys of these sub-sampled units involved the measurement of all physical parameters outlined in Form 4 Technical Circular No.8 (Johnston and Slaney, 1996). The survey was conducted by a crew of three people over a 10-day period between September 5th and September 14th, 1997.

The physical parameters outlined in Form 4 Technical Circular No.8, pg. 92-93 (Johnston and Slaney, 1996) were measured in a two-stage process. Initially, while the crew walked the stream, one crew member hip-chained sample sites and recorded the length of each hydraulic unit. Individual habitat units were classified according to definitions outlined in Technical Circular No.8 (Johnston and Slaney, 1996) as pools, glides, riffles and cascades. The measurements of habitat units provided information on the length and the frequency at which individual habitat units occurred throughout the reach. Observations were recorded regarding side channel length, frequency and their potential for restoration, the height and passability of waterfalls, slope or bank failures and signs of habitat degradation. Photographs of representative sections of each reach, migration barriers and degraded habitat units were also taken. These can be found in Appendix A.

Secondly, when a habitat unit needed to be measured in detail, all physical parameters were measured rather than visually estimated. Methodologies pertaining to the measurement of physical parameters of fish habitat are described in detail in Technical Circular No.8 (Johnston and Slaney, 1996). The various parameters assessed in each habitat unit and their method of measurement are shown in Table 2. All habitat data was entered into Form 4 of Technical Circular No.8 (Johnston and Slaney, 1996) to ensure consistency in data collection. These can be found in Appendix B.

Table 2. Physical parameters assessed for each habitat unit within a sample site.

Confirmation of the presence or absence of salmonids within the Clisbako River and McFarland Creek sub-basins was determined by electrofishing. Selected habitat units were randomly sampled using the single-pass removal method. All fish captured were identified, assessed for life stage and recorded in Form 5 of Technical Circular No.8 (Johnston and Slaney, 1996). If the identity of any fish was in doubt, the taxonomic keys in Field Key to the Freshwater Fishes of British Columbia (McPhail and Carveth, 1993) and/or Fresh Water Fishes of Canada (Scott and Crossman, 1990) were used. Detailed methods of the electrofishing techniques used in this survey are outlined in the Fish Stream Identification Guide Book, July 1995 and the Lake and Stream Inventory, Standards and Procedures (RIC, 1995).

The assessment of fish habitat in the Clisbako River and McFarland Creek sub-basins was based on the calculation of various habitat parameters. By comparing these values to set diagnostic values, fish habitat was rated as "good", "fair" or "poor". The diagnostic values used to compare assessed habitat values, as well as a detailed habitat evaluation procedure, can be found in Table 5 of Technical Circular No. 8 (Johnston and Slaney, 1996). The criteria outlined in Table 5 were used except for certain habitat parameters listed below:

• Percent Pools

• Off-Channel Habitat

• Holding Pools

• Access to Spawning Areas

Also, Appendix E, Questions for Habitat Evaluation, from Technical Circular No.8 (Johnston and Slaney, 1996), was used to identify potentially degraded or limiting salmonid habitats. Values and ratings were calculated for: percent pools, pool frequency, number of LWD pieces per channel width, percent cover, dominant and subdominant substrates, off-channel habitat, spawning gravel quantity/quality, access for spawning adults and number of adult holding pools. Redd scour was not evaluated due to a lack of historical information on spawning beds. Values for all parameters mentioned above have been entered into Form 6 of Technical Circular No.8 (Johnston and Slaney, 1996).

This section describes the findings of the fish habitat assessment and the fish sampling completed in Clisbako River and McFarland Creek. The field data collected during these assessments can be found in Appendix B. Presented in Table 3 is a summary of these characteristics by reach. A total of 634 individual habitat units were observed throughout the mainstem of the Clisbako River; 41% (257 units) of these were assessed in detail. Presented in Table 4 are the habitat unit frequencies, average channel characteristics within specified reaches and overall reach lengths.

Table 3. Summary of channel characteristics for the Clisbako River by reach.

Table 4. Level 1 Field Assessment - Habitat Diagnosis Summary Form

The Clisbako River was divided into reaches based on the homogeneity of stream gradient, channel morphology and substrate. The Clisbako River mainstem was divided into 10 reaches up to a 15m falls approximately 24.8km upstream of the mouth. Reach boundaries were based on the findings of the Level 1 Fish Habitat Assessment, which provide a more detailed view of the stream. Consequently, these boundaries do not correspond with those suggested in the Overview report. Reach 1 of McFarland Creek, a tributary to the Clisbako River was also assessed.

Reach 1 of the Clisbako River was 717m long, with an average gradient of 1.0%. Eleven habitat units from a total of 28 units, were sampled in detail in this reach. The channel was relatively unconfined on the right bank, and confined on the left bank by a steep gully wall, which had become the fill slope for a road. The channel was a riffle-bar-pool type with habitat diversity being moderately distributed between glides (53%), riffles (35%) and pools (12%). Representative photographs of the three habitat types are in Appendix A (photos 1, 2, 3). Reach 1 had an average bankfull width of 12.5m and an average wetted width of 9.0m. The mean water depth was 0.39m and the mean residual pool depth was 0.89m. The dominant and subdominant substrates were gravel and cobble, and compaction and siltation were high. Boulder cover (42%) and overhanging vegetation (26%) dominated in-stream cover. There was also one LWD jam that had created a high quality pool that was providing good cover; this type of habitat was limited in this reach. This was the only holding pool deep enough to offer cover for fish in the reach. There was one 5m long sidechannel in this reach; however, it had poor access as only 2m at the downstream end would be accessible to fish at low flows. The only disturbance indicators noted in this reach were an LWD jam and siltation of the substrates.

Large woody debris (LWD) was moderately abundant in this reach with an extrapolated rating of 2.3 (good). This value indicates that there were 2.3 pieces of LWD per average bankfull width. However the majority of the LWD in this reach was located in one large, opened debris jam approximately 315m upstream from the Nazko River confluence (Appendix A, photo 4). Other than this jam there was very little LWD in the remainder of the reach. The LWD that was present was small to medium sized and found along the edges of the channel.

Percent pools in this reach is rated as "poor" with a value of 12%. Pool frequency, which assesses the distance between pool habitats, was also rated as "poor" with a value of 19.9 bankfull widths between pools. This value indicates that there are approximately 244m between pools. The percent wood cover in pools was rated as "poor" with a value of 10%. Boulder cover in gravel-cobble riffles was rated as "good" with a value of 41.3%, while overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "good" with a value of 26.2%. Boulder cover (42%) and overhanging vegetation (26%) dominated in-stream cover. Substrates in reach 1 were dominated by gravels and sub-dominated by cobble. However, both boulders and fines were present throughout the reach. Substrate type and quantity was rated as "good", however, due to the presence of silts and fines and the high compaction of the substrate the quality of the substrate was rated as "poor". Off-channel habitat, a measure of the number of off-channel units per km of stream, was rated as "fair" with a value of 1.4.

Holding pools, which function as resting areas during upstream migration, were rated as "good", with a value of 4.2 pools per km of stream. There was only one pool that fit the definition of a holding pool, however, and because this pool was 37m long, the results are misleading. Access to spawning areas was rated as "good", although there was a beaver dam approximately 315m upstream from the Nazko River confluence (Appendix A, photo 4). This dam may act as an obstacle to migration.

Spawning gravel quantity in this reach was rated as "good" due to the size of the dominant substrate; however, spawning gravel quality was rated as "poor" due to the presence of silts and fines. These silts and fines have filled the interstitial spaces in the gravel, which has resulted in the high compaction of the substrate. This high compacted substrate does not allow for easy building of redds or for proper gas exchange in incubating eggs.

No major valley wall or channel stability problems were noted in this reach. However, an access road parallels the left side of this reach for its entire length. In the middle sections, the fill slope of this road is actually the valley wall and is contributing unknown amounts of sediment to the creek. Also, this area is used for free range cattle grazing and sections of the stream banks have been damaged due to cattle crossings and are contributing sediment.

Shrub/herb and young mixed forest vegetation dominated the riparian zone, however, some mature conifer trees are present within the reach. The canopy closure was relatively poor (0-20%) and did not afford much thermal protection to the stream. The riparian zone in certain areas of this reach was limited to a narrow band of shrubs and/or individual trees. Beyond the riparian area on the right side of the stream and the initial 300m of the left side was grazing land used by cattle.

Reach 2 of the Clisbako mainstem was 835m long with a gradient of 2%. Seven habitat units from a total of 13, were sampled in detail. The channel was confined and was a riffle-pool type. Habitat diversity was considered low with 74% riffle, 23% glide, and 3% pool. Representative photographs of this reach are presented in Appendix A (photos 5, 6, 7). Reach 2 had an average bankfull width of 13.3m and an average wetted width of 10.0m. The mean water depth was 0.29m and the mean residual pool depth was 0.83m. Cobbles were the dominant substrate within this reach, with gravel and boulders being equally subdominant. The compaction and siltation of the substrate was low, increasing the quality of the available spawning habitat for anadromous salmon and resident trout. There were 28 pieces of LWD in this reach, but only six were functional. Boulder cover (37%) and overhanging vegetation (22%) dominated in-stream cover. There was one pool in reach 2 with a residual depth of 0.83m, and consequently was too shallow to be considered a holding pool. There was no off-channel habitat observed in this reach. The main disturbance indicators in this reach were low pool frequency, extended riffle zones and debris parallel to the bank; however many small, boulder controlled tertiary pools were present, which provide good cover and improve the habitat.

Large woody debris (LWD) was moderately abundant in this reach with an extrapolated rating of 1.6 (fair). This value indicates that there were 1.6 pieces of LWD per average bankfull width. However the majority of the LWD in this reach was located along the edges of the channel and was of a small size class. There was no large debris in this reach spanning the width of the channel.

Percent pools in this reach is rated as "poor" with a value of 3%. Pool frequency, which assesses the distance between pool habitats, was also rated as "poor" with a value of 61 bankfull widths between pools. This value indicates that there are approximately 811m between pools. The percent wood cover in pools was rated as "poor" with a value of 0%. Boulder cover in gravel-cobble riffles was rated as "good" with a value of 37%, while overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "good" with a value of 22%. Boulder cover and overhanging vegetation dominated in-stream cover. Substrates in reach 2 were dominated by cobble and sub-dominated by equal shares of gravel and boulder. Substrate type, quantity and quality were rated as "good". The substrate was not compacted and did not have a significant amount of fines, thus the interstitial spaces in the gravel and cobble, and the boulder cover, were useable for rearing and over-wintering. There was no off-channel habitat in this reach.

There were no holding pools in this reach, which gave a "poor" rating for this characteristic. Access to spawning areas was rated as "good" because there were no obstructions downstream of this reach.

Spawning gravel quality in this reach was rated as "good" due to the size the dominant substrate and the absence of fines; however, spawning gravel quantity was rated as "poor" as spawning gravel were only noted in small pockets throughout the reach.

No major valley wall or channel stability problems were noted in this reach.

A mixed deciduous/coniferous mature forest dominated the riparian area in this reach. The canopy closure was poor (0-20%).

Reach 3 of the Clisbako mainstem was 389m long with a gradient of 4%. Six of the ten habitat units were sampled in detail. The creek is characterized by a confined channel with 52% riffle, 25% pool, 19% cascade, and 4% glide. Reach 3 had an average bankfull width of 11.1m and an average wetted width of 9.3m. The mean water depth was 0.54m and the mean residual pool depth was 0.98m. Representative photographs of this reach are in Appendix A (photos 8, 9, 10). The dominant and subdominant substrates were boulder and cobble respectively. Compaction was high due to the large size of the substrate. For the most part, the substrate in this reach was too large to provide good spawning habitat. There were 17 pieces of LWD in this reach, but only three were functional. Boulders were the primary source of cover (45%), with two holding pools providing secondary cover (32%). No off-channel habitat was observed in this reach. The disturbance indicators noted in this reach, although minor, were elevated mid-channel bars, eroding banks and debris parallel to the bank.

Large woody debris (LWD) was moderately abundant in this reach with an extrapolated rating of 1 piece of LWD per bankfull width (fair). However the majority of the LWD in this reach was located along the edges of the channel and was of the small size class. There was no large debris spanning the width of the channel in this reach.

Percent pools in this reach is rated as "poor" with a value of 25%. Pool frequency was also rated as "poor" with a value of 4.7 bankfull widths between pools. This value indicates that there is approximately 52m between pools. There was no wood cover in pools. Boulder cover in gravel-cobble riffles was rated as "good" with a value of 45%. Overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was rated as "good" with a value of 32%. Boulder cover and deep pools dominated in-stream cover. Substrates in reach 3 were dominated by boulder and sub-dominated by cobble. Substrate type, quantity and quality were rated as "good". The substrate was compacted but did not have a significant amount of fines, thus the interstitial spaces in the gravel and cobble and the boulder cover was useable for rearing and over-wintering. There was no off-channel habitat in this reach.

Holding pools were rated as "good", with a value of 5.1 pools per km of stream. Access to spawning areas was rated as "good" because there were no obstructions downstream of this reach.

Spawning gravel quality in this reach was rated as "good"; however, spawning gravel quantity was rated as "poor" as spawning gravel was only noted in small pockets throughout the reach.

No major valley wall or channel stability problems were noted in this reach. However, some minor bank erosion was apparent.

A mature coniferous forest dominated the riparian zone in this reach. The canopy closure was low (0-20%).

Reach 4 of the Clisbako mainstem was 2,203m long with a gradient of 0.5%. Of the 95 habitat units in this reach, 35 were sampled in detail. The channel varied between confined and frequently confined, and was a riffle-bar-pool type. The habitat diversity was moderately distributed between riffles (44%), glides (33%), pools (22%), and cascades (1%). The average bankfull width was 12.3m with an average wetted width of 9.0m. Representative photographs of this reach are in Appendix A (photos 11, 12, 13, 14, 15). The mean water depth was 0.35m and the mean residual pool depth was 1.07m. The dominant and subdominant substrates were gravel and cobble, but small amounts of sand and boulder were present. Compaction was moderate, which may have decreased the quality of spawning habitat where it was available. Boulder and deep pool provided the primary cover, with five holding pools providing additional cover for adult salmon. There were no off-channel habitats in this reach. The disturbance indicators noted in this reach were extensive areas of unvegetated bar, eroding banks, large extensive sediment wedges and elevated mid-channel bars.

Large woody debris (LWD) was moderately abundant in this reach with an extrapolated rating of 1.5 pieces of LWD per bankfull width (fair). The majority of the LWD in this reach was located in three debris jams (Appendix A, photos 16, 17), other pieces of LWD were noted scattered through other habitat units. The LWD was small to medium sized. The LWD was both in and across the channel, providing good cover and acting as hydraulic controls.

Percent pools in this reach is rated as "poor" with a value of 22%. Pool frequency was also rated as "poor" with a value of 8.6 bankfull widths between pools. This value indicates that there are approximately 106m between pools. The percent wood cover in pools was rated as "poor" with a value of 1.4%. Boulder cover in gravel-cobble riffles was rated as "fair" with a value of 10%, while overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "fair" with a value of 10%. Boulder cover and deep pools provided the majority of the in-stream cover. Substrates in reach 4 were dominated by gravels and sub-dominated by cobble. However, both boulders and sand were present as subdominant substrates in several of the habitat units in the reach. Substrate type was rated as "fair", due to the presence of sand as a subdominant in a few of the habitat units. There was no off-channel habitat in this reach. The abundance boulders coupled with frequent functional LWD and clean cobble and gravel suggests that this reach provides a moderate amount of summer rearing and overwintering habitat.

Holding pools were rated as "good", with a value of 3.2 pools per km of stream. Access to spawning areas was rated as "good" because there were no obstructions downstream of this reach.

Spawning gravel quantity and quality in this reach was rated as "good" for both anadromous and resident species.

No major valley wall or channel stability problems were noted in this reach. However, small sections of eroding bank and mid-channel bars were noted.

The riparian zone in this reach was mature conifer forest, with a deciduous/coniferous, shrub and grass species understorey. Deciduous shrubs are present along most of the channel length and provide some overhanging bank vegetation cover. The canopy closure was low throughout this reach (0-20%). There has been no logging near or adjacent to the creek channel in this reach.

Reach 5 was 835m long with an average gradient of 1.5%. Five habitat units from a total of 12, were sampled in detail. The channel was confined and of the riffle-pool type with habitat diversity being distributed between riffles (54%), pools (34%), glides (8%) and cascades (4%). The average bankfull width was 13.0m, with an average wetted width of 9.7m. Representative photographs of the reach are in Appendix A (photo 18). The mean water depth was 0.3m; no residual depth was recorded because the only pool in the reach was not sampled in detail. The dominant and subdominant substrates were cobble and boulder and the compaction was high; this reduced the quantity and quality of the available spawning habitat. Boulders provide the majority of in-stream cover in this reach. There were no holding pools assessed in this reach and there were no off-channel habitats. The disturbance indicators noted in this reach were elevated mid-channel bars, multiple channels, extensive riffle zones and eroding banks.

Large woody debris (LWD) was abundant in this reach with an extrapolated rating of 3.8 (good). This value indicates that there were 3.8 pieces of LWD per average bankfull width. However, most of the functional LWD in this reach was located near the downstream end of the reach in a large jam (Appendix A, photo 19) Few pieces of functional LWD were noted in other habitat units. The LWD that was present was small to medium sized with a few pieces falling into the large category. The mature timber in this area is naturally small.

Percent pools in this reach is rated as "poor" with a value of 34%. Pool frequency was also rated as "poor" with a value of 42 bankfull widths between pools. This value indicates that there are approximately 546m between pools. The percent wood cover in pools was not rated, as the only pool in the reach was not sampled in detail. Overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "good" with a value of 39%; boulder cover (30%) and overhanging vegetation provided the majority of this cover. Substrates in reach 5 were dominated by cobbles and sub-dominated by boulders and rated as "good"; however, the substrates were heavily compacted. There was no off-channel habitat in this reach. The abundance boulders, and associated tertiary pools, coupled with overhanging vegetation and one large pool suggests that this reach provides a moderate amount of summer rearing and over-wintering habitat. However, rearing and over-wintering habitats could be limited to some extent by the high compaction of the substrate.

Holding pools were rated as "good", with a value of 1.2 pools per km of stream. Access to spawning areas was rated as "good" because there were no obstructions downstream of this reach.

Spawning gravel quantity in this reach was rated as "poor" for both anadromous and resident species, due to the size the dominant substrates and the lack of suitably sized gravel. Spawning gravel quality was rated as "poor" due to the high compaction. The highly compacted substrate does not allow for easy building of redds or for proper gas exchange in incubating eggs.

No major valley wall or channel stability problems were noted in this reach. However, small sections of eroding bank and mid-channel bars were noted. A natural cobble, boulder talus slope was present in this reach and has entered the channel (Appendix A; photo 20).

The riparian zone in this reach is a mixed mature forest with 0-20% canopy cover. Deciduous shrubs are present along most of the channel length and provide some overhanging bank vegetation cover. There has been no logging near or adjacent to the creek channel in this reach.

Reach 6 of the Clisbako mainstem was 4,246m long with a gradient of 0.5%. Thirty-nine habitat units from a total of 110 units were sampled in detail. The channel was occasionally confined and was of the riffle-bar-pool type with habitat diversity being evenly distributed between pools (36%), riffles (35%), and glides (29%). The average bankfull width was 12.5m with an average wetted width of 8.9m. The mean water depth was 0.34m and the mean residual pool depth was 1.46m. Representative photographs of the reach are in Appendix A (photos 21, 22, 23, 24, 25, 26, 27). The dominant and subdominant substrates were gravel and cobble with some sand. Overall compaction was low and siltation was limited to areas of slower water; consequently, pockets of spawning gravel remained free of silt. Deep pools provide the primary cover. The beaver activity increased substantially from previous reaches, and their dams were creating numerous pools. There were 11 off-channel habitat units, with nine of these being side- or back channels; seven had good access and two had poor access. The other two off-channel habitats were relic channels and had poor access. The disturbance indicators noted in this reach were eroding banks, extensive areas of bar, elevated mid-channel bars, recently formed debris jams, LWD parallel to the bank and extensive riffle zones.

Large woody debris (LWD) was moderately abundant in this reach with an extrapolated ratingof 2 (good). This value indicates that there were 2 pieces of LWD per average bankfull width. However, most of the functional LWD was located in five LWD jams (Appendix A, photos 28, 29, 30); one of these LWD jams may make passage difficult for large adult salmon, but it is not a barrier. Only 23 other pieces of functional LWD were noted in other habitat units. The LWD that was present was small to medium sized. Debris cover in pools was rated as "poor".

Percent pools in this reach is rated as "poor" with a value of 36%. Pool frequency was also rated as "poor" with a value of 12.6 bankfull widths between pools. This value indicates that there are approximately 158m between pools. The percent wood cover in pools was rated as "poor" with a value of 2%. Boulder cover in gravel-cobble riffles was rated as "poor" with a value 1.5%, while overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "poor" with a value of 2.3%. Deep pools and beaver ponds provided the majority of the cover in this reach. Substrates in reach 6 were rated as "good", dominated by gravel and sub-dominated by cobble. Off-channel habitat, a measure of the number off-channel units per km of stream, was rated as "fair" with a value of 2.8. A total of 11 off-channel habitats, totaling 406m in length were noted, with nine of these being side- or back channels; seven had good access and two had poor access. The other two off-channel habitats were relic channels and had poor access. The abundance of pools and beaver ponds in this reach coupled with the cover provided in the interstitial spaces of the bed material provide an adequate amount of summer rearing and over-wintering habitat. However, rearing and over-wintering habitats could be limited to some extent by the lack overhead and boulder cover in the reach.

Holding pools were rated as "good", with a value of 4.3 pools per km of stream. Access to spawning areas was rated as "good"; however the presence of many beaver dams and natural debris jams in the reach may impede migration. Two obstructions were noted that would make passage difficult for large adult salmonids. The first was a beaver dam with a rise of 1.0m; the second obstruction was a LWD jam.

Spawning gravel quantity in this reach was rated as "good" for both anadromous and resident species. However, spawning gravel quality was rated as "poor", due to the presence of large quantities of sand in several of the habitat units. It should be noted that the sand was localized and that good quality, uncompacted spawning gravel was available in many of the habitat units in the reach.

No major valley wall or channel stability problems were noted in this reach. However, small sections of eroding bank and mid-channel bars were noted. Also, a logging road parallels this reach for a distance and its fill slope may be inputting sediment to the stream. The riparian zone between the creek and the fill slope of the road consists of shrub and alder (Appendix A, photo 31). Natural gully wall raveling was also contributing minor amounts of sediment into the stream (Appendix A, photo 32).

A mature coniferous forest with a deciduous shrub understorey dominated the riparian zone. The canopy closure of this reach is relatively poor (0-20%). Logging has occurred along this reach, and there is a logging road within the riparian area. It appears that the width of the leave strip adjacent to the creek is sufficient to buffer any influence from the cutblock.

Reach 7 of the Clisbako mainstem was 15,352m long with a gradient of 1.0%. One hundred and forty-six individual habitat units, from a total of 345 units, were sampled in detail. The channel varied between confined and frequently confined and was a riffle-pool type with habitat diversity being moderately distributed between riffles (56%), pools (24%), glides (19%) and cascades (1%). The average bankfull width was 11.7m with an average wetted width of 8.7m. The mean water depth was 0.33m and the mean residual pool depth was 0.79m. Representative photographs of this reach are in Appendix A (photos; 33 - 40). The dominant and subdominant substrates were cobble and gravel with some boulders. The compaction was moderate, which may reduce the quality of spawning gravel available. Boulders, LWD and deep pools provided the primary cover, with 17 holding pools (greater than 1m residual depth) creating additional cover for spawning fish. The beaver activity remained high in this reach as beaver dams created 15 of the pools. There were ten off-channel habitats in this reach; one was a relic channel with poor access, the remaining nine were side- and backchannels, where eight had good access and one had poor access. The disturbance indicators noted in this reach were extensive areas of unvegetated bar, multiple channels, elevated mid-channel bars, LWD parallel to the banks, extensive riffle zones and eroding banks.

Large woody debris (LWD) was moderately abundant in this reach with an extrapolated ratingof 2 (good). This value indicates that there were 2 pieces of LWD per average bankfull width. However, most of the functional LWD was concentrated in 17 LWD jams (Appendix A, photos 41 - 45). Also, 14 habitat units each had more than 15 pieces of LWD. The LWD that was present was small to medium sized. Debris cover in pools was rated as "fair". Many of the debris jams provided fish cover and acted as hydraulic controls within the reach.

Percent pools in this reach is rated as "poor" with a value of 24%. Pool frequency was also rated as "poor" with a value of 14.5 bankfull widths between pools. This value indicates that there are approximately 170m between pools. The percent wood cover in pools was rated as "fair" with a value of 6.6%. Boulder cover in gravel-cobble riffles was rated as "fair" with a value 19%, while overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "poor" with a value of 6.7%. Deep pools, beaver ponds, boulders and LWD provided the majority of the cover in this reach. Substrates in reach 7 were dominated by cobble and sub-dominated by gravel and were rated as "fair", due to its moderate compaction. Off-channel habitat, a measure of the number off-channel units per km of stream, was rated as "fair" with a value of 1. A total of 10 off-channel habitats, totaling 1410m in length were noted, with nine of these being side- or back channels; eight had good access and one had poor access. The remaining off-habitat was a relic channel. The abundance of pools and beaver ponds in this reach coupled with the cover provided by boulders and LWD provide an adequate amount of summer rearing and over-wintering habitat. However, rearing and over-wintering habitats could be limited to some extent by the moderate compaction of the substrates in the reach.

Holding pools were rated as "good", with a value of 2.5 pools per km of stream. Access to spawning areas was rated as "good"; however the presence of many beaver dams and natural debris jams in the reach may slow migration. There were two beaver dams that obstructed fish migration; one dam was 1.5m high with a water depth at the outfall of 0.05m, the second dam was 1.5m high with a water depth at the outfall of 0.15m.

Spawning gravel quantity in this reach was rated as "poor" for both anadromous and resident species, due to the large size of the substrate. However, the spawning gravel quality was rated as "good". It should be noted that the bed material was assessed to be moderately compacted, which may limit the ability of fish to spawn in the available spawning gravel.

No major valley wall or channel stability problems were noted in this reach. However, small sections of eroding bank and mid-channel bars were noted (Appendix A; photo 46). A natural gully wall with a talus slope at its base may be contributing a minor amount of coarse rock substrate to this reach (Appendix A; photo 47). A shrub/alder riparian area does exist between the talus slope and the stream channel.

A mature mixed forest, with a deciduous shrub understorey dominated the riparian zone. The canopy closure of this reach is relatively low (0-20%). Deciduous shrubs and/or grasses are present along most of the stream banks in this reach. These shrubs and grasses do provide some cover within the reach. No timber harvesting has taken place near or adjacent to the creek channel in this reach.

Reach 8 of the Clisbako mainstem was 335m long with a gradient of 6.0%. Three individual habitat units, from a total of nine, were sampled in detail. The channel was confined and was a riffle-pool type with habitat diversity being distributed between riffles (68%), pools (20%), cascades (9%) and glides (3%). The channel width averaged 10.5m and the average wetted width was 7.1m. The mean water depth was 0.35m and the mean residual pool depth was 0.68m. Representative photographs of this reach are in Appendix A (photo 48). The dominant and subdominant substrates were boulder and cobble. Compaction was high. There was little LWD present, and boulders provided the majority of the cover in this reach. Two off-channel habitats were noted. The disturbance indicators observed were elevated mid-channel bars, multiple channels and extensive riffle zones.

Large woody debris (LWD) was scarce in this reach with an extrapolated rating of 0.69 (poor). This value indicates that there were 0.69 pieces of LWD per average bankfull width. The LWD that was present was in the small size category and was found along the edges of the creek providing very little cover or hydraulic control. Debris cover in pools was rated as "poor".

Percent pools in this reach is rated as "fair" with a value of 20%. Pool frequency was also rated as "fair" with a value of 7.8 bankfull widths between pools. This value indicates that there are approximately 82m between pools. The percent wood cover in pools was rated as "poor" with a value of 0%. Boulder cover in gravel-cobble riffles was rated as "fair" with a value 26.5%, while overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "good" with a value of 35%. Boulders provided the majority of the cover in this reach. Substrates in reach 8 were dominated by boulder and sub-dominated by cobbles and were rated as "good". Off-channel habitat, a measure of the number off-channel units per km of stream, was rated as "good" with a value of 9. Two side channels, totaling 170m in length were noted. The abundance of boulder cover in this reach coupled with the tertiary pools associated with them provide good summer rearing and over-wintering habitat.

Holding pools were rated as "good", with a value of 9 pools per km of stream. Access to spawning areas was rated as "good" because there were no obstructions downstream of this reach.

Spawning gravel quantity in this reach was rated as "poor" for both anadromous and resident species, due to the large size of the substrate. However, the available spawning gravel quality was rated as "good".

No major valley wall or channel stability problems were noted in this reach.

A mature coniferous forest with low canopy closure (0-20%) dominates the riparian zone in this reach. No timber harvesting has taken place near or adjacent to the creek channel in this reach.

Reach 9 of the Clisbako mainstem was 185m long with a gradient of 12%. Two individual habitat units, from a total of seven, were sampled in detail. The channel was entrenched and was of the block-step-pool type with habitat diversity being moderately distributed between cascades (51%) cascade, riffles (25%), and pools (24%). The average bankfull width was 11.8m with an average wetted width of 8.52m. The mean water depth was 0.64m and the mean residual pool depth was 1.13m. Representative photographs of this reach are in Appendix A (photos 49, 50, 51). The dominant and subdominant substrates were bedrock and boulder. Compaction was high. There were 18 LWD pieces in this reach. Boulder and deep pool provided the primary cover. One off-channel habitat exists in this reach. Only one disturbance indicator, multiple channels, was noted in this reach.

Large woody debris (LWD) was abundant in this reach with an extrapolated rating of 6.6 (good). This value indicates that there were 6.6 pieces of LWD per average bankfull width. However, of the LWD present only five pieces were functional and were of the small and moderate size class. Debris cover in pools was rated as "poor".

Percent pools in this reach is rated as "fair" with a value of 24%. Pool frequency was also rated as "fair" with a value of 3 bankfull widths between pools. This value indicates that there are approximately 35m between pools. The percent wood cover in pools was rated as "poor" with a value of 0%. The only boulder cover in this reach was in cascade habitat, not gravel-cobble riffles and consequently was not assessed; however, overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "fair" with a value of 22.5%. Boulders and deep pools provided the majority of the cover in this reach. Substrates in reach 9 were dominated by bedrock and sub-dominated by boulders and were rated as "poor". Off-channel habitat, a measure of the number off-channel units per km of stream, was rated as "good" with a value of 5.4. One side channel, 25m in length were noted. The abundance of boulder cover in this reach coupled with the deep pools provides some summer rearing and over-wintering habitat. Over-wintering habitat is likely limited in this reach due to the absence of other cover types, especially the lack of substrate interstitial space.

Holding pools were rated as "good", with a value of 10.8 pools per km of stream. Access to spawning areas was rated as "poor" as there were several sets of cascades, which would pose obstacles to upstream migration.

Spawning gravel quantity and quality in this reach was rated as "poor" for both anadromous and resident species, due to the bedrock - boulder substrate.

No major valley wall or channel stability problems were noted in this reach.

A mature coniferous forest dominates the riparian zone in this reach. The canopy closure was poor (0-20%). No timber harvesting has taken place near or adjacent to the creek channel in this reach.

Reach 10 of the Clisbako mainstem was 176m long with a gradient of 12%. One out of the five habitat units in this reach was sampled in detail. The channel was entrenched and of the boulder-step-pool type with habitat diversity distributed between cascades (62%), riffles (30%), and pools (8%). The average bankfull width was 11.7m with an average wetted width of 7.8m. The mean water depth was 0.34m and the mean residual pool depth was 0.68m. The dominant and subdominant substrates were boulder and cobble. There were 15 pieces of LWD within the reach. Boulder cover provided the primary cover. There were no off-channel habitats or disturbance indicators noted in this reach.

Large woody debris (LWD) was abundant in this reach with an extrapolated rating of 3.4 (good). This value indicates that there were 3.4 pieces of LWD per average bankfull width. However, of the LWD present, only 2 pieces were functional and were of the small and moderate size class.

Percent pools in this reach was rated as "poor" with a value of 8%. Pool frequency was also rated as "poor" with a value of 15 bankfull widths between pools. This value indicates that there are approximately 176m between pools. The percent wood cover in pools was rated as "poor" with a value of 0%. Boulder cover in gravel-cobble riffles was not assessed as this condition did not exist in the habitat units that were assessed in detail; however, overhead cover, comprised of LWD, boulders, cutbanks, and overhanging vegetation, was also rated as "good" with a value of 25%. Boulders provided the majority of the cover in this area. Substrates in reach 10 were dominated by boulder and sub-dominated by cobble and were rated as "good". There was no off-channel habitat in this reach.

Holding pools were rated as "good", with a value of 6.9 pools per km of stream. Access to spawning areas was rated as "poor" as there was a 15m falls, a migration barrier, at the upstream reach break (Appendix A: photo 52). It was not possible to obtain UTM data at this point.

Spawning gravel quantity and quality in this reach was rated as "poor" for both anadromous and resident species, due to the high boulder component in the substrate.

No major valley wall or channel stability problems were noted in this reach.

A mature coniferous forest dominates the riparian zone in this reach. The canopy closure was poor (0-20%). No timber harvesting has taken place near or adjacent to the creek channel in this reach.

Habitat diagnosis values for McFarland Creek were not calculated because the gradient exceeded levels that allow for feasible rehabilitation. Also, the channel type was boulder-step-pool, which has low salmonid habitat value (Johnston and Slaney, 1996).

Reach 1 of McFarland Creek was 507m long with an average gradient of 10%. Seventeen habitat units were observed in this reach. The channel was entrenched and was of the boulder step pool type. The average bankfull width was 3.3m, while the average wetted width was 1.8m. The dominant and subdominant substrates were boulder and cobble respectively. Cover was dominated by boulders, while very little LWD was present. The riparian area was composed of mature conifer forest. The canopy closure was low (0-20%). Extensive logging has taken place adjacent to the non-fish bearing reaches immediately upstream reach 1. Representative photographs of this reach are located in Appendix A (photos 53-56). A 6m waterfall over rough bedrock is located 507m upstream from the confluence with the Clisbako River is a barrier to fish migration. At the time of the survey, only a trickle of water was flowing over the falls (Appendix A, photo 57). The gully walls in the upper portion of reach 1 were very unstable and have naturally slid into the channel (Appendix A, photo 58). However, it is not expected that this material reached the Clisbako mainstem as there was little evidence to suggest that the material moved from its initial deposition area. The bed material indicates that flows in McFarland Creek are likely insufficient to move large amounts of sediment.

Habitat diagnosis values for McFarland Creek were not calculated because the gradient exceeded levels that allow for feasible rehabilitation. The channel type was identified as boulder step pool, which typically has low salmonid habitat values (Johnston and Slaney, 1996). This precludes the feasibility of any habitat rehabilitation work in this reach.

The main fish bearing area impacted by timber harvesting is reach one of the Clisbako River mainstem. Road development has introduced fine sediment from roads, surface erosion and fill slopes into the system. Agricultural use has resulted in the erosion of stream banks. Sediment from both these sources has infilled the substrate interstitial spaces with an accumulation of fines and sands. The increase in the amount of fine sediment being input into the stream has resulted in heavily compacted bed material. Control structures, such as LWD, are relatively uncommon, which has reduced the complexity of the stream and has resulted in the lack of pools found in this reach. Also, the substrate has been compacted through the accumulation of fines to a point where substrate scouring, caused by control structures, is limited.

Rainbow trout are resident throughout the accessible reaches of the Clisbako River and its tributaries, while chinook salmon and mountain whitefish seem to be limited to the lower reaches of the system. If the impacts discussed above encroach on the habitats used by the various life stages of these species, a reduction in productivity will likely result. The condition of the fish habitat in this system has been assessed and rated (section 4.0) using the diagnostic criteria outlined in Technical Circular No. 8 (Johnston and Slaney, 1996). These ratings are based on general habitat conditions developed for coastal streams and may not necessarily be indicative of this interior watershed. Professional judgement was used to determine if the calculated ratings were real or exaggerated. In this case, the upper seven reaches (3-9) of the Clisbako River appear to be unimpacted by past logging in the area. These reaches flow through a steeply entrenched valley, which remains unharvested and any cutblocks in the area are well back from the creek, above the topographic break of the valley. For these reasons, it has been assumed that these reaches are, for the most part, in their natural state and that the diagnostic values calculated in the results can be used as a baseline to compare the impacted areas in other reaches. For example, the amount of pool habitat in reaches 3 through 9 averaged 26%, which is a poor rating according to the diagnostic table. However, the amount of pool habitat in reaches 1 and 2 was only 11% and 3%, respectively. These values would indicate that the pool habitat in reaches 1 and 2 has been severely impacted. This lack of pools in reach 2 is due to a channel type of boulder riffle. In this type of channel, pool habitat is limited to small pools forming below the boulders. These small pools do not meet the minimum pool size requirement in Technical Circular No. 8 (Johnston and Slaney, 1996) and were not counted because they were too small. This minimum size requirement may not be valid in small streams or boulder dominated reaches, where small pools make up a significant portion of the available habitat. Reaches 2 and 10 were considered to be unimpacted and in their natural state. They were not included in the calculation of average channel characteristics because their boulder riffle and cascade channel types are not typical of the Clisbako sub-basin. For the most part, the diagnostic ratings indicate problems or limiting habitats; however, this is suspected to be the natural state of the habitat. Reach 1 was compared to these average characteristics for reaches 3 through 9 (Table 5), as these reaches offer a more accurate representation of the natural state of the in-stream habitat in this watershed.

Table 5. Comparison of the impacts in reach 1 versus the natural averaged characteristics of reaches 3-9.

Stream edges and protected pools with areas of low velocity and shallow water depths usually provide good summer rearing habitat for rainbow trout fry (Burt, 1996). Boulder and cobble provide ideal cover in pools; however, LWD, undercut banks and other cover types will be utilized if they are available (Burt, 1996). The presence of cobble and boulder averages "good" to "fair in the system, but our results indicate that pool habitat (% pools) is lacking. This finding suggests that summer fry habitat is limited, thus fry production may be affected and likely not optimal. The scarcity of pools is partially due to the inherent nature of the system; however, the amount of pool habitat in reach 1 (12%), even when compared to the control reaches (26%), is obviously a limiting factor to fry production in this reach. This deficiency is possibly a result of natural deposition, or past infilling of existing pools by sediment transported downstream from reach based sediment sources (roads and cattle crossings). It is magnified by the inability of the system to scour new pools due to the high compaction of the substrate and the lack of control structures.

Chinook juveniles usually migrate to sea as fry, before the winter of the first growing season; however, some populations are known to stay in fresh water for up to a year (Scott and Crossman, 1990). In the Clisbako system, chinook fry migrate from the Nazko River into the lower reaches of the Clisbako mainstem to rear before they migrate to sea. While rearing in fresh water, chinook parr utilize pool, pool tail-out and edge habitat. Rainbow trout parr, on the other hand, utilize the deeper and higher velocity section of the stream for summer rearing. Boulder and white water are the preferred forms of cover, yet other types of available cover will be used. Boulder glides and riffles, as well as pools with moderate velocities, provide excellent habitat for this life stage. The habitat conditions in reach 1 are not conducive to high chinook or rainbow trout parr productivity. Despite these conditions, rainbow trout parr productivity may be at an acceptable level system- wide due to the natural state and quality of the rearing habitat upstream of reach 1.

Juvenile rainbow trout and chinook salmon either move into the interstitial spaces of the substrate or migrate into deep pools or larger rivers during the winter months. Stream beds of established cobble and boulder substrate with minimal infilling of fines provide the ideal overwintering habitat. The availability of cobble or boulder in most of the reaches indicates that there is an acceptable amount of overwintering habitat in the system; however, it was noted in reach 1 that fines are accumulating in the channel and are infilling the interstices causing the substrate to compact, which results in the degradation of overwintering habitat. Off-channel habitat, a source of overwintering habitat for parr, is also limited within reach 1.

It is assumed that a portion of the resident fish migrate into the Nazko River for the winter months, which may reduce the amount of in-stream habitat needed for the existing population.

Adult salmon require deep, cool holding pools during upstream migration. This assessment shows that there were few good holding pools for adult chinook in the system. This deficiency of deep pools is likely due to the inherent channel characteristics in the system. Spawning resident trout do not require as many deep, high quality holding pools, thus the scarcity of deep holding pools affects these spawning fish to a lesser extent, yet it does effect the amount of overwintering habitat available to adult trout.

Rainbow trout and mountain whitefish spawn in loose gravel and cobble ranging from 0.6 to 10.2 cm in size (Whyte et al., 1996; Bjornn and Reiser, 1991; Scott and Crossman, 1990). Spawning gravel in the reaches assessed were determined to be present in moderate to high quantities and of moderate to high quality, with the exception of reaches 1 and 2 which had a larger substrate base and only intermittent sections of small to medium size gravel and cobble. Chinook salmon on the other hand use these larger substrates (large gravel and cobble) for spawning. The substrates in reaches 1 and 2 were considered to be the appropriate size for chinook spawning, but the substrate in reach 1 was too compacted and possibly too silty to allow for spawning and survival of the eggs.

Incubation success seriously diminishes when excessive amounts of fines are present in the substrates (Whyte et al., 1996). Oxygen supply to the eggs and waste removal from the eggs is obstructed in such cases. In reach 1 fines are accumulating in the channel and it is likely that some infilling of the interstices is occurring, impairing the gas exchange efficiency of the substrate.

In summary, the major limitations to fish production in this system appear to be the lack of rearing habitat and holding areas in reach 1. The infilling of interstitial spaces with fines and the resulting compaction of the substrate is also a concern in spawning areas and overwintering substrates.

In-stream habitat restoration of the Clisbako River should be directed at areas revealing the most habitat damage or limitations and which presently, or historically, are utilized by anadromous and resident fish species. This restoration effort should be limited to areas where logging, agricultural and other man-influenced perturbations exist.

This study found that overall, the majority of the lower 10 reaches of the Clisbako mainstem, that are accessible to fish, were relatively unimpacted by unnatural processes and were in a natural state. Nevertheless, habitat impairment in reach 1 of this system was found. The major problems are the lack of pools and control structures and the high compaction and siltation of the substrate.

In reach 1 the silt is suspected to originate from the mainline road which runs close and parallel to the stream for the entire length of the reach and eventually crosses the mainstem at the upper reach boundary. In certain areas along the road, the steep fill slope of the road, and the road crossing itself, are likely contributing sediment to river. Also the land surrounding this reach is used for grazing cattle. The movements of the cattle are contributing to the sediment input to the stream. This sediment input has probably resulted in the filling of interstitial spaces in the substrate, which has caused the substrate to be compacted. The limited number of pools is likely a result of the lack of control structures in the reach and the reduced ability of the stream to scour the substrate in its compacted state. Control structures such as LWD will typically cause scouring of the substrate and the eventual creation of pools. Furthermore, protective cover of all types is generally scarce in this area. Functional LWD and boulders, which form pools and provide cover are scarce. All these factors have effectively limited the rearing and spawning habitat available in this reach.

The problems described above warrant the following restoration activities for the Clisbako River watershed:

Instream restoration should focus on activities that will increase the overall productivity of the stocks inhabiting this watershed. To do this, restoration techniques which increase the amount or availability of juvenile rearing habitat should be used. Techniques such as the introduction of LWD and boulder clusters to the stream channel initiate the formation of scour pools, create cover for juvenile fish, and introduce additional hydraulic control to the stream, which will slow or reduce the stream’s energy and generally increase the productivity potential of the watershed.

• Lack of pool and overwintering habitat.
• Compacted bed material.

• Increase the amount of LWD or boulder clusters.
• Build a fence along the banks of reach 1 to limit the impact on the channel caused by cattle fording the creek.
• Control sediment sources using cribwalls, or re-vegetate the slope.

• Habitat capabilities in reach 1 of the Clisbako River appear to be limited by a low abundance of pool areas, and compacted bed material. However, these limitations appear to be the result of logging and agricultural activities in the watershed. It is imperative to address up-slope problems before commencing any in-stream rehabilitation.
• Place LWD in reach 1 to increase pool habitat and cause bed material scouring and cleaning. Excavate initial pools downstream of the LWD.
• Fencing of reach 1 is another option to limit the amount of damage caused by agricultural use (e.g. cattle crossing). This may not be feasible because the cattle would likely find another area to cross the stream.

The feasibility of completing rehabilitation work on any stream is largely affected by the cost, which in turn is affected by the availability of materials and the access to the site. The access to the stream channel within reach 1 of the Clisbako River is relatively good as a logging road comes within 100m of the stream edge in the upper portion, and agricultural access is available in the lower section of the reach. Both adequate size boulders and LWD are available onsite in the upper part of the reach.

Another alternative to the above-recommended prescriptions to improve rearing habitat in the reach would be to excavate a rearing pond or side channel within the reach. There is a flat, open field easily accessible by heavy equipment at the top end of the reach to the left of the bridge crossing. This site would need to be surveyed in detail before a design plan could be developed. Information required would include elevation profiles and a hydrologic assessment of the site. The elevation profiles should include the channel adjacent to the proposed development site as well as contour profiles encompassing the banks and floodplain areas.

It is felt that in-stream placement of LWD would be the more cost effective prescription, as there is a chance that the scour effect caused by increasing the amount of LWD or control structures may not only increase pool habitat, but also loosen and clean the compacted bed material, resulting in improved spawning gravel.

Costs for rehabilitation work in Clisbako River will be provided once more information and detail regarding the extent and feasibility of prescriptions is acquired for individual sites from the Level 2 assessment.

C. von Hahn, pers. comm. to L. Chew of Carmanah Research Ltd., 1996.

Appendix A. Photographs of significant habitat features and reach-representative sites

Appendix B. Level 1 - Habitat Survey Data Forms

1 Field Copies

2 Digital Print-outs

Appendix C. Copy of Field Notes

Appendix D. Fish Distribution Data Form