Prepared for

West Fraser Mills Ltd.

P.O. Box 6000

Quesnel BC, V2L 3J5

by

R. P. Griffith and T. Roy

Carmanah Research Ltd.

Building 4, 203 Harbour Road

Victoria BC, V9A 3S2

Telephone: (250) 920-9900 Facsimile: (250) 920-9800

Email: fish @carmanah.com Http:\\www.carmanah.com

March 1998

Carmanah Research Ltd. is conducting studies in the Cottonwood River drainage, near Quesnel BC, to identify and restore fish habitat degraded by forest harvesting, road building, and mining. Initial fish habitat assessments within the Sovereign, Victoria, and Reddish Creek sub-basins identified the following stream sections as being of greatest concern:

• Reaches 4 to 6 of Fontaine Creek, tributary to Reddish Creek;
• Reaches 1 to 5 of Sovereign Creek;
• Reach 1 of Chipp Creek, tributary to Sovereign Creek.

To evaluate the status of fish habitat within these stream sections and assess rehabilitation potential and needs, field investigations were conducted in October 1997. The site-specific concerns identified in the Overview were inspected at various locations within the Victoria, Fontaine and Sovereign Creek systems.

A severe and universal lack of pool habitat, both in abundance and frequency, reflects channel aggradation in all stream sections addressed in the Level 1 FHAP assessments. In addition to infilling with bed materials, the lack of pool habitat appears to be greatly attributable to the loss of stable large woody debris (LWD). The lack of natural replacement of LWD is due to extensive deforestation of riparian areas. LWD installations would appear to be the most appropriate instream activity to mitigate such losses and restore the abundance and frequency of pool habitat in all stream sections surveyed in 1997.

Due to the presence of large beaver impoundments in Reach 4 and the lower section of Reach 5 of Fontaine Creek, pool abundance is much higher. In flowing channel sections within these reaches, however, fish habitat conditions are consistent with those described above. The same applies to Reach 6, where significant beaver activity was not observed. The beaver ponds in Reaches 4 and 5 may provide the opportunity to greatly increase cover for fish through large additions of LWD. However, this should not be viewed as an alternative to the rehabilitation of fluvial pool frequency and abundance through LWD installations in flowing stream sections.

The sites for LWD installations have been tentatively identified. However, for all stream sections surveyed, the dominance of small streambed and bank materials, coupled with diverse and abundant indications of channel disturbances, raises concerns about stability and the success and durability of rehabilitation activities. In Sovereign Creek, the frequency and extent of unstable banks suggests that instream installations might be detrimental by contributing to instability. In this system, initial rehabilitation activities might best be focused on promoting bank stability wherever necessary and appropriate. In Fontaine Creek, conditions appear to be more stable. All final selection and evaluation of installation sites must await qualified geomorphic and engineering evaluation. In order to obtain such input (and in keeping with WRP protocol and guidelines), Level 2 FHAP assessments are prescribed for all of the stream sections addressed in the 1997 Level 1 assessments. Level 1 and Level 2 FHAP assessments should also be considered for sections of Horan Creek, a tributary to Fontaine Creek. Eskridge Creek, a tributary to Sovereign Creek, should also be considered. Forest harvesting has disturbed both tributaries. They were not evaluated under this contract because the disturbances noted during 1997 field investigations were not readily apparent at the overview level.

A historical placer mining site on Sovereign Creek, just upstream of Eskridge Creek, should also receive immediate engineering evaluation geared towards stabilization of severe erosion. Measures should be taken to stabilize and vegetate the exposed and eroding banks remaining at historical placer mining sites adjacent to Sovereign Creek and Fontaine Creek. As a lower priority, engineering evaluation should also be obtained for the installation of instream structures to disrupt laminar flows and hopefully promote scour pools within a long riffle section at the top of Fontaine Creek Reach 1.

In Victoria Creek, two debris jams in Reach 1 have the potential to form total barriers to fish passage. They should be removed, either selectively or entirely. Other jams inspected in this stream should be monitored to ensure continued passage of fish. In a small unnamed tributary to Reach 2 of Victoria Creek, improvement of a culvert is not warranted due to lack of habitat potential upstream. Below the culvert a 200 m section of the channel (found to contain both chinook and rainbow juveniles) has been severely impacted by clearcutting, and warrants immediate attention. As a first measure, the extensive unstable debris that is choking this section should be selectively removed so that further rehabilitation needs can be determined. Unstable debris should also be selectively removed in the lowermost 1 km of the stream draining the Robertson Lake system, tributary to the top of Victoria Creek Reach 3. These materials may contribute to the formation of jams in Reaches 1 to 3 of Victoria Creek, and also form barriers to fish passage within the tributary itself.

All interpretation of findings and formulation of recommendations are based solely on the condition of habitat for fish, with consideration of species (and associated requirements) reportedly present in different stream sections. However, quantitative biological assessments to enable projection of potential increases in fish production have not been undertaken.

The authors would like to acknowledge and thank the many people who contributed to this report. The project was managed and directed by Carmanah Fisheries Biologist Kevin Brydges. Field work was carried out by Fisheries Biologist Thomas Roy with the assistance of Fisheries Technicians Henrik Liebe-McGinnis and Rob Hollingshead. Mr. Al Hunter, TFL Forester with West Fraser Mills Ltd., provided guidance and logistical support. Mike Parker, WRP Specialist, provided valuable technical input to the study design. Carmanah Geographic Information Systems (GIS) specialists A. Dewey and M. Burrell prepared maps and figures.

H. Liebe-McGinnis contributed to quality assurance and data compilation.

Executive Summary *

Acknowledgements *

Table of Contents *

List of Tables *

List of Figures *

List of Appendices *

1.0 Introduction *

2.0 Watershed Information *

3.0 Summary of Overview FHAP Findings and Recommendations *

4.0 General Considerations *

5.0 Methods *

6.0 Results (1997 Investigations) *

7.0 Discussion and Recommendations *

References *

Table 1. Physical parameters assessed for habitat units within a sample site. *

The Cottonwood River drainage is located 18 km north of Quesnel, British Columbia, to the east of the Fraser River into which it empties. On behalf of West Fraser Mills Ltd., Quesnel Division, Carmanah Research Ltd. (Carmanah) completed a Level 1 Fish Habitat Assessment Procedure (FHAP) of three sub-basins within the Cottonwood River watershed. These are the Sovereign, Victoria, and Reddish Creek sub-basins (Figure 1). Based on the findings in the Overview, the Sovereign, Victoria, and Reddish creek sub-basins were surveyed for negative effects on fish habitat that are related to forestry activities. R.P. Griffith and Associates, who prepared this document on behalf of Carmanah, interpreted the results of the field investigations, completed in October 1997.

Under the B.C. Watershed Restoration Program (WRP), the Overview FHAP identifies the following:

1) areas of concern with respect to impacts on fish habitat within developed watersheds;

2) general assessment of rehabilitation needs and potential; and

The following synopsis is from the Overview FHAP of the Cottonwood River watershed (Carmanah, 1997).

The Cottonwood River flows west into the Fraser River from the Interior Plateau of central British Columbia (Figure 1). The joining of the Swift River and Lightning Creek forms the Cottonwood River. The Cottonwood River watershed drains an area of 247,363 ha from a portion of the Interior Plateau called the Quesnel Highlands. The average elevation increases towards the eastern boundary of the watershed. The eastern boundary of the watershed abuts the Willow and Bowron River watersheds, while the southern and western boundaries abut the Quesnel River watershed.

The sub-basins surveyed in this Level 1 FHAP report, the Sovereign, Victoria and Reddish creeks, flow into the Swift River.

Settlements within the Cottonwood River watershed form a population base of less than 500 people. Hamlets in the watershed include Cottonwood, Wingdam and Cinema. Highway 97 passes over sections of the Cottonwood River. Highway 26, which provides access to Wells, Barkerville and Bowron Lake Provincial Park, follows Lightning Creek before crossing into the Willow Creek watershed. Transmission lines, underground pipeline right-of-ways and railways also pass through the Cottonwood River watershed.

Forestry and mining are the main economic endeavours in the Cottonwood River watershed. The watershed lies within the boundaries of the Quesnel and Prince George Forest Districts. Forest harvesting has occurred in all sub-basins. The major operating timber licensees in the watershed are West Fraser Mills Ltd., Weldwood of Canada Ltd. and Dunkley Lumber Ltd. The provincial Small Business Forest Enterprise Program also operates in the watershed. Placer and hydraulic mining has taken place in the watershed since 1860. This has been identified as having a significant impact on the stream systems in the Cottonwood River watershed (AIM, 1996; Rowland and McDonald, 1996).

Mining within the watershed has occurred for the past 100 years and continues today. There are upwards of 100 mineral and placer claims within the watershed boundaries. The majority of historic mining activity has taken the form of placer gold mining of paleochannels. The abundant glacial and glaciofluvial materials have yielded many million ounces of gold. Some underground mining has also occurred. The Lightning and John Boyd Creek sub-basins contain the most concentrated areas of mining-related disturbances in the Cottonwood system (Chapman and Dobson, 1997). Placer mining requires the disturbance of surficial materials and has resulted in obvious and significant impacts on both the material constitution and stability of creek channels. Hydraulic mining has also resulted in surficial disturbance. The levels of resultant sedimentation and consequent channel changes remain undetermined.

Two biogeoclimatic zones dominate the study area. These zones are the Sub-Boreal Spruce zone (SBS) and the Engelmann Spruce-Subalpine Fir zone (ESSF). The SBS zone is found in the low to mid-elevation portions of the study area. It has a continental climate characterized by warm, moist summers with long growing seasons and cool winters (Meidinger and Pojar, 1991). Mature forest containing the hybrid white spruce and subalpine fir covers much of the SBS landscape. The ESSF zone occupies the mid- to upper elevations of the Cottonwood River watershed. As it borders on the lower elevation SBS zone, it has similar but cooler climate. A distinctive feature of the ESSF landscape is extensive young and maturing forests of subalpine fir mixed with subalpine meadows. There is a small portion of Alpine Tundra zone in the high elevations along the eastern boundary of the Cottonwood River watershed.

The Cottonwood River watershed is affected by continental climates. Most of the annual precipitation falls as snow in the winter months. Summer rainstorms are typically of short duration with high intensity showers. The Water Survey of Canada operates four gauging stations in the watershed for measuring seasonal changes in water flow. These are the Swift Creek gauging station 08KE003, Cinema gauging station 08KE009, Wingdam gauging station 08KE004 and Little Swift River gauging station 08KE024. Mean annual flow for the watershed is 26.0 m³ s^-1 (Rowland and McDonald, 1996). Average maximum flow takes place during April and May with flows measured at 234.0 m³ s^-1. Average minimum or low water flow periods occur during December through March. Summertime and wintertime 7-day low water flows are 4.15 m³ s^-1 in September and 3.51 m³ s^-1 in December to January respectively.

The Cottonwood River watershed provides migration, spawning, incubation and rearing habitat for rainbow trout (Oncorhynchus mykiss), coho (O. kisutch), pink (O. gorbuscha) and chinook salmon (O. tshawytscha) (Fish Information Stream Summary (FISS), 1994). Chinook salmon are present in the lower reaches of the Swift River, Sovereign and Victoria creeks (Imhof and Sutherland, 1995). A waterfall on the Swift River may limit migration of chinook salmon (Rowland and McDonald, 1996). Bull trout (Salvelinus confluentus) and Dolly Varden (S. malma) are distributed throughout the Swift and Little Swift rivers and Lightning Creek (FISS, 1994). Mountain whitefish (Prosopium williamsoni) have also been noted in the Swift River and Victoria Creek (Carmanah, 1996).

Forest Renewal British Columbia (FRBC) has funded several projects directed toward improving the breadth and quality of information for the Cottonwood River watershed. Fish inventories on the Victoria, Sovereign and Reddish creek sub-basins improved assessments of fish habitat by providing an interpretation of habitat impacts. Carmanah Research Ltd. (1996) completed the fish inventory reports. Other Watershed Restoration projects in the Cottonwood River watershed include the following:

• Sediment Source Survey Level 1 (Carmanah Research Ltd., in progress),
• Integrated Watershed Restoration Plan (Carmanah, in progress),
• Riparian and Fish Habitat Overview Level 1 (Carmanah, in progress),
• Fish and Fish Habitat Inventory (Carmanah, in progress),
• Sediment Source Inventory and Mapping (CARR Ecological Consultants Ltd. and AIM Ecological Consultants Ltd., 1996),
• Riparian, Wetland and Terrestrial Assessment (AIM Ecological Consultants Ltd., 1996) and
• Level 1 Interior Watershed Assessment Procedure (IWAP) (Dobson Engineering Ltd., 1995).

The following summary of findings and implications is from the draft Overview FHAP report (Carmanah, 1997).

Victoria Creek (Figure 2) is approximately 54 km in length, and has an average stream gradient of approximately 0.4%. It contains chinook salmon, rainbow trout and bull trout, burbot, mountain whitefish, lake chub, longnose suckers and redside shiners. The most widely distributed salmonid appears to be rainbow trout, reportedly occurring throughout the mainstem and most tributaries. Bull trout are reported to occur in the Victoria Creek mainstem and some larger tributaries, such as Chiaz Creek. Chinook are reported to spawn as far upstream as Reach 5 (approx. 22 km), but juveniles have been reported in tributaries to Reaches 7 and 8, a further 14 to 15 km upstream (FHAP Map # 93A-071, -072, -081, 93B-090, -100).

Six percent of the Victoria Creek sub-basin has been harvested. The system has a low level of concern regarding forestry-related disturbances within the entire Cottonwood River drainage. However, logjams were identified on the mainstem and recommended for field evaluation of origins and implications with respect to fish passage. One culvert crossing on a small tributary to the Victoria Creek mainstem was also recommended for closer evaluation, as it was an area of impacted streambank associated with a homestead site.

Figure 2. Victoria Creek drains an area of 29,949 ha as it flows west into the Swift River.

Reddish Creek is 10 km in length, and has a mean gradient of approximately 0.2% (Figure 3). The major tributary is Fontaine Creek (FHAP Map # 93A-091, -092), which is approximately 12 km in length and has an average gradient close to 2.7%. Rainbow trout are the only salmonids reported in the Reddish Creek sub-basin. They are present throughout the system, including Fontaine Creek and its tributaries.

Forest harvesting has been extensive within the Reddish Creek drainages. Previous assessments of this sub-basin have indicated high risk factors associated with peak flows, surface erosion and the condition of riparian buffers. In terms of impacts, Reddish Creek itself is relatively well buffered by its low gradient and its extensive associated wetlands. There are various impacts on the wetlands, but no evidence of channel changes or other degradation of the stream itself. The same does not apply to Fontaine Creek.

The first kilometer of Fontaine Creek is flat and bordered by wetlands, but the gradient increases progressively upstream. Reaches 4 to 6 (total 8.2 km), which extend to the headwaters (FHAP Map # 93A-092), have been subjected to substantial forest harvesting, including deforestation of the riparian vegetation (on one bank and/or the other) over much of this stream length. Negative impacts were predicted from the Overview and a Level 1 FHAP was recommended for these reaches. Similar impacts were noted on Horan Creek, a major tributary of Fontaine Creek, which flows into the creek at Reach 5 (FHAP Map # 93A-092). A channelized section below a bridge in Reach 1 of Fontaine Creek was also recommended for further evaluation, as were two sites of earlier placer mining in Reach 2 and a culvert crossing on an unnamed tributary at the top of Fontaine Creek Reach 5.

Figure 3. Reddish Creek drains an area of 7,245 ha as it flows south into the Swift River

Sovereign Creek is 25 km long, and has a mean gradient of approximately 2.7% (Figure 4). This sub-basin has also been subjected to extensive forest harvesting, and risks associated with peak flows, surface erosion and disruption of riparian buffers have been assessed as high. Sedimentation of streambeds has been identified as a particular problem within this sub-basin. A total of 73 specific sediment sources have been determined, of which 59% are attributable to forest harvesting activities. Five particularly high-risk sites are related to earlier placer mining development.

Forestry-related impacts on the creek are the clearcutting and disturbance of the riparian vegetation along the Sovereign Creek mainstem. The lower 8.7 km of the stream (Reaches 1 and 2) were identified as areas with potential negative effects on fish habitat. However, substantial impacts were observed as far as Reach 5, a further 6.8 km upstream (FHAP Map # 93A-091).

Chinook salmon can access Reach 3 of Sovereign Creek, but are believed to be able to access Reach 4. These fish are also reported to utilize Chipp Creek, which is tributary to the top of Sovereign Creek Reach 2 (FHAP Map # 93A-091). As in Fontaine Creek, rainbow trout are reported to occur throughout the sub-basin, with the exception of the uppermost 1.5 km of headwaters (Reach 10).

A Level 1 FHAP was recommended for Reaches 1 to 5 of the Sovereign Creek mainstem. Five other specific locations were identified for separate evaluation, including an isolated oxbow and a side channel area on the Sovereign Creek mainstem, an old mill site and two earlier placer mining sites. Due to substantial impacts of forest harvesting on the riparian zone of the lower 1.5 km of Chipp Creek, this stream section was also recommended for Level 1 assessment. A Level 1 FHAP was recommended for upper Eskridge Creek, another tributary to Sovereign Creek located just upstream of Reach 5 (FHAP Map # 93A-091, 93H-001). Eskridge Creek contains rainbow trout.

Figure 4. Sovereign Creek drains an area of 11,249 ha as it flows west into the Swift River.

For the Sovereign, Victoria, and Reddish Creek sub-basins of the Cottonwood River drainage, the identification of fish species and their distribution, as reported here, is based on existing records (Carmanah, 1997). Additional fish sampling was conducted at some locations, in this Level 1 assessment, as warranted by specific circumstances (e.g. fish absence/presence above potential obstructions). However, such sampling was not in any way quantitative.

In the absence of intensive investigations of fish populations themselves, it can rarely be said with a high level of confidence that the fish habitat in any given system is inadequate to support either present or historic (i.e. pre-development) levels of fish production. Nor can it be said with a high level of confidence that any habitat rehabilitation activities will result in increased fish production within a system, even though such activities may result in the desired habitat improvements.

All discussions and recommendations presented here relate solely to an evaluation of the habitat itself, based on the underlying philosophy of the FHAP that if fish are present, their production will increase with rehabilitation of disturbed habitat (P. Slaney, pers. comm.).

Although the FHAP does not address the size or dynamics of fish populations, it does require identification of target species (Johnston and Slaney, op. cit.). For the Sovereign, Victoria, and Reddish Creek sub-basins, these species and their documented distribution have been identified. Salmonids are the principal target species of the FHAP, in this case chinook salmon, rainbow trout and bull trout.

The habitat conditions and requirements are based on the specific needs and preferences of these species. In addition to spawning opportunities, this also includes consideration of specific microhabitats used by juveniles of each target species. For chinook salmon and rainbow trout, various guidelines are provided within the WRP Technical Circular No. 9 (Fish Habitat Rehabilitation Procedures) and a host of other documents (e.g. extensive discussions and bibliography provided by Griffith (1980)). The presence and distribution of bull trout in British Columbia has recently been identified (Haas and McPhail, 1991). Concerted efforts are being made to assemble all available data in order to determine the specific habitat requirements and preferences of these fish in B.C. systems (BC Fisheries Branch, in prep.). McPhail and Baxter (1992), Reiman and McIntyre (1993), and Hemmingsen and Buchanan (1994) have compiled a preliminary reviews on bull trout habitat requirements. These references are consistent with the author’s experience with bull trout in river systems within the B.C. Interior (e.g. Griffith, 1995; 1997).

3. Special Consideration of Bull Trout

The recent identification and research of bull trout warrants specific mention, particularly in the context of forestry-related impacts on stream habitats. Bull trout are believed to be particularly sensitive to forestry-related disturbances. Channel and hydrologic instability, sedimentation and aggradation of bed materials, the reduction of large woody debris (LWD) and other complex cover, elevated stream temperatures and the blockage of fish migrations (McPhail and Baxter, 1992.; Reiman and McIntyre, 1993; Hemmingsen and Buchanan, 1994.).

The patchy distribution of bull trout populations has been noted both between and within systems (McPhail and Baxter, 1992.; Reiman and McIntyre, 1993; Hemmingsen and Buchanan, 1994.). Bull trout exhibit extensive and complex migratory behaviour both between and within systems indicating that this species may have more demanding habitat requirements than other salmonids. For instance, the period of egg incubation for bull trout is generally extended. This increases susceptibility of the eggs to the effects of sedimentation and bedload movement. Juvenile bull trout exhibit a strong preference for pool habitat, a habitat that may diminish as a result of infilling with sediments and aggradation of bed materials.

Juvenile densities appear to be highly correlated with the abundance of complex LWD cover, especially in pools and other quiet habitats. (This was extremely evident in studies of bull trout in the Yalakom River and Kwoiek Creek drainages in the southwest Interior of B.C. (Griffith, 1995, 1997). Reductions in such cover generally accompany forest harvesting, so populations may be expected to decline (Reiman and McIntyre, 1993.). Juvenile bull trout may utilize other forms of cover in the absence of LWD, but these may also be subject to infilling as a result of forest harvesting (Reiman and McIntyre, 1993.).

The migratory behaviour of bull trout is thought to be attributable to highly specific habitat requirements at different life stages. Water temperature may be of key importance, and appears to be the most consistent factor influencing the distribution of this species. Migrations may also occur between summer feeding areas and winter refuge areas. Migrations may cover any portion of a given stream, and may also extend into other streams and lakes. Consequently, the maintenance of migration corridors is of utmost importance to this species. The maintenance of migratory patterns may be vital to the genetic integrity of any given population (Reiman and McIntyre, 1993).

The presence of bull trout may be very hard to detect as populations are often small and isolated. Due to their benthic orientation, the efficiency in sampling for bull trout is lower than that for other salmonids (e.g. Griffith, 1997). Accordingly, in a system where any presence of bull trout is detected, the possibility of a widespread distribution should be acknowledged. With the documentation of bull trout in Victoria Creek, this caution should be applied to the entire Cottonwood River drainage.

 

4. Instream Habitat Rehabilitation Techniques

Approaches and guidelines provided in the WRP Technical Circular No. 9 (Fish Habitat Rehabilitation Procedures) have been the principal references for considerations and recommendations presented here. Most citations refer to various contributors to this publication. Other guidelines and specifications developed for the improvement of fish habitat in streams have also been consulted (Anon., 1986; 1989).

This document does not profess to include a full appreciation of fluvial geomorphology or hydrology. The principal focus in following discussions is biological, based on specific needs and preferences of relevant species within given stream sections. Recommendations are essentially based on most appropriate biological objectives, applied to generic guidelines provided in the documents identified above. In keeping with FHAP procedures (Johnston and Slaney, 1996) and WRP rehabilitation guidelines (e.g. Cederholm et.al., 1997), recommendations also include consideration of natural templates observed in the field.

However, as emphasized by various contributors to the WRP guidelines (Allan and Lowe, 1997; Cederholm et.al., op. cit.; Slaney and Martin, 1997), all of the following must be conducted (or overseen) by a fully qualified and experienced fluvial morphologist or hydraulic engineer: 1) verification (or not) of proposed techniques and associated feasibility; 2) identification of specific installation sites, as required (e.g. LWD additions); 3) determination of precise numbers and densities of restorative undertakings, consistent with the latter; and 4) full site-specific evaluation of potential durability, implications with respect to channel and bank stability, and ultimate probabilities of attaining the desired biophysical results. For the systems addressed here, it is assumed that such input will take the form of a Level 2 FHAP. This is in keeping with established protocol (Johnston and Slaney, op. cit.).

Contributors to the WRP fish habitat rehabilitation guidelines also emphasize that instream rehabilitation may be ineffective and short-lived if not accompanied by simultaneous rehabilitation of upslope areas and riparian zones (e.g. Cederholm et.al., op. cit.; Slaney and Martin, 1997). Stabilization of the watershed, or at least "storm proofing," is often a necessary prerequisite to successful fish habitat rehabilitation, in medium-sized streams at least (Allan and Lowe, 1997).

It is assumed that rehabilitation of upslope areas and riparian zones will precede or accompany any instream rehabilitation activities in the Cottonwood River watershed. In particular, rehabilitation of all disrupted riparian zones should be initiated and completed as a top priority. This should involve assured presence of coniferous species, even where deciduous second growth is well advanced (Cederholm et.al., op. cit.). It is further assumed that areas of present bank instability will be addressed as a priority, and that the avoidance of further destabilization of such areas will be addressed in the ultimate siting and specifications for any and all instream works, including additions of LWD (ibid.).

Fish habitat assessments were completed for reaches within the Sovereign Creek and Reddish Creek sub-basins as well as specific sites in Victoria Creek that had been identified as high priority in the OFHAP. These included Reaches 1 to 5 of Sovereign Creek, Reach 1 of Chipp Creek, and Reaches 4 to 6 of Fontaine Creek. The methods used followed those outlined in the Fish Habitat Assessment Procedures manual (WRP Technical Circular No. 8; Johnston and Slaney, 1996).

Reach boundaries were based on the findings of the OFHAP and verified, when possible, during the field surveys. The Sovereign Creek mainstem was divided into 11 reaches, Reddish Creek into 3 reaches, Fontaine Creek into 7 reaches and Chipp Creek into 5 reaches.

The assessed reaches were sampled using a subsampling fraction of 1/3 for each habitat unit. This fraction guaranteed a detailed analysis of at least 33.3% for all pools, glides, riffles and cascades. On occasion, some habitat units were subsampled at 1/2 or 1/1 when stream conditions warranted a higher fraction. Conversely, when habitat units were found to be very short and numerous the subsampling fraction was adjusted to 1/5. Thus, habitat unit averages were extrapolated over the entire reach from our detailed assessments. The surveys of these subsampled units involved the measurement of all physical parameters outlined in Form 4 of Tech. Circ. 8. This survey was conducted during October 2-19, 1997.

The physical parameters outlined in Form 4 were measured in a two-stage process. Initially, while the crew walked the stream, one member hip-chained the sample site and recorded the length of individual habitat units. This chaining process provided information on the length and frequency at which individual habitat units occurred throughout the reach. During this process, notations were made of side channels and their length and accessibility, the height and passability of waterfalls, presence of slope or bank failures and other signs of habitat degradation. Photographs of subsampled units, representative sections of each reach, migration barriers and degraded habitat units were also taken (Appendix 3). Photodocumentation procedures followed those outlined in A Guide To Photodocumentation (RIC, 1996).

Secondly, when a habitat unit needed to be measured in detail (subsampled), all physical parameters were measured, not visually estimated (Table 1). All habitat data were entered into waterproof field copies of Form 4 to ensure consistency in data collection. Habitat data are presented in Appendix 2.

To aid in mapping stream features and disturbances, UTM co-ordinates were obtained for locations of significant tributaries, reach breaks, slumps etc. Co-ordinates were gathered and stored in a Trimble GeoExplorer II handheld GPS unit, referenced in a field notebook and later downloaded onto computer disc. After the field assessment the data were corrected and used by Carmanah’s Geographic Information Systems (GIS) technicians to generate FHAP Maps (referred to throughout the report) of the study area showing the position of each site.

Table 1. Physical parameters assessed for habitat units within a sample site.

Confirmation of the presence or absence of salmonids at specific sites within the Victoria Creek sub-basin was determined by electrofishing. A single pass removal method was performed in riffle, run and pool habitat in both disturbed and undisturbed sections of creek.

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 referenced. Detailed methods of electrofishing are outlined in the Fish Stream Identification Guidebook, (Anon., 1995) and the Lake and Stream Inventory, Standards and Procedures (RIC, 1995).

The evaluation of fish habitat in both the Sovereign and Reddish sub-basins was based on the calculation of various habitat parameter statistics. By comparing these values to set diagnostic values, fish habitat was rated as "good," "fair" or "poor." The criteria outlined in Table 5 were used except for the habitat parameters listed below:

• Offchannel Habitat

• Holding Pools

• Access to Spawning Areas

Appendix E of Tech. Circ. 8, Questions for Habitat Evaluation, was used to identify potentially degraded or limiting salmonid habitats within the surveyed reaches. All diagnostic values were derived from measurements of primary habitat units, which had been measured for length and width over the entire reach. Habitat parameters included calculations for percent pools, pool frequency, number of LWD pieces per bankfull channel width, percent cover, dominant and subdominant substrates, offchannel 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 Tech. Circ. 8 and are presented in Appendix 4.

Photographs referred to in the sections below are bound separately in Appendix 3.

The Level 1 FHAP assessments on Fontaine, Sovereign and Chipp creeks constituted the focus of the 1997 field survey. Site assessments were also conducted on Eskridge Creek and other small unnamed tributaries. (See supplied maps for locations of specific site assessments.)

The results and implications for each stream surveyed will be addressed separately in following sections. At the outset, however, it is useful to compare all results collectively. The results for all surveyed sections are summarized in Table 2 (Fontaine Creek Reaches 4 to 6), Table 3 (Sovereign Creek Reaches 1 to 3) and Table 4 (Sovereign Creek Reaches 4 and 5, and Chipp Creek Reach 1). Indicators of the occurrence of extensive riffles and the limited frequency and extent of pools have been omitted from these tables, as these and other aspects of general macrohabitat are dealt with in greater detail in Tables 5 to 7 (corresponding to Tables 2 to 4 above).

The most informative results are the rates of observed disturbances, expressed as occurrences (data records) per kilometer of stream length. In all cases, the most frequent disturbances are eroding banks, braiding and/or elevated mid-channel bars.

Table 2. Recorded disturbance indicators for habitat surveys in Fontaine Creek, Reaches 4 to 6; October 1997.

Table 3. Recorded disturbance indicators for habitat surveys in Sovereign Creek, Reaches 1 to 3; October 1997.

Table 4. Recorded disturbance indicators for habitat surveys in Sovereign Creek, Reaches 4 and 5, and Reach 1 of Chipp Creek; October 1997.

Table 5. General summary of habitat composition in surveyed sections of Fontaine Creek, Reaches 4 to 6; October 1997.

Table 6. General summary of habitat composition in surveyed sections of Sovereign Creek, Reaches 1 to 3; October 1997.

Table 7. General summary of habitat composition in surveyed sections of Sovereign Creek, Reaches 4 and 5, and Reach 1 of Chipp Creek; October 1997.

All of these factors indicate aggradation of the stream channels surveyed (Hogan and Ward, 1997). The prevailing channel type is riffle-pool, dominated by gravel and/or cobble bed materials (RPg-w or RPc-w; ibid.), which reflects the relatively low gradient of most sections surveyed, particularly those on Sovereign Creek (Tables 6 to 8). Due to steeper gradients, cobble cascade-pool (CPc-w) morphology is dominant in Reach 6 of Fontaine Creek and Reach 1 of Chipp Creek (ibid.).

In channels of these types, where existing bed materials are small, the most profound effect of aggradation is the process of smoothing the streambed (Cederholm et.al., 1997) as pools and other deeper habitats become infilled. The result is a progressive increase in the extent of shallow riffles and glides. This is clearly evidenced for all stream sections surveyed in 1997 (Tables 5 to 7).

Although Reach 1 of Fontaine Creek may appear to be an exception, this is only due to the influence of beavers. As will be detailed in the following section, the great majority of pool habitat in this reach is due to beaver impoundments. Implications with respect to aggradation in flowing channel sections are the same here as they are elsewhere.

Consistent with findings of the Overview (Carmanah, 1997), the highest rates of disturbances are in Sovereign Creek, especially Reach 1 (Table 3). The proportion of pool habitat is particularly low within this reach (6% of wetted area; Table 6). Generally speaking, it is in the lowermost sections of streams that the effects of aggradation may be most evident as signs of diminishing gradient and cumulative deposits from upstream sources (Hogan and Ward, 1997).

In this context, it is interesting to note that the rate of disturbances in Reach 4B of Sovereign Creek is second only to that in Reach 1 (Table 4). Furthermore, the proportion of pool habitat in Reach 4B (2.5% of wetted area) is the lowest overall (Table 8), and riffles are most abundant (74% of wetted area). This is despite considerably higher gradient (avg. 2.5%) than Reach 1 (1%) and other sections of the Sovereign Creek mainstem that were surveyed (Tables 7 and 8).

Reach 5 was considered as the natural template for the 1997 investigations. Although this reach has suffered disturbances from forest harvesting in adjacent areas and/or further upstream (Table 5), it has retained a greater diversity of macrohabitat than any other stream section surveyed (Table 8). In addition to geographic considerations, this is likely related to the maintenance of mature riparian buffers along much of the stream length subjected to forest harvesting in this reach (ibid.).

Braiding and elevated mid-channel bars are the most frequent disturbance indicators recorded for all three reaches surveyed on Fontaine Creek (Table 3). There is some occurrence of bank instability (Photo 1), but this is neither severe nor widespread (Table 3). Braiding was noted fairly frequently in Reaches 4 and 5, but it is generally attributable to the effects of the numerous beaver dams on these reaches. Some of these dams might constitute barriers to the upstream passage of small resident rainbow trout (Photo 2), but they also account for much of the pool habitat in Reach 4. Pool area is far more abundant here (56%) than in any other stream section surveyed (Tables 6 to 8).

It seems likely that such impoundments are historic phenomena, as evidenced by an extensive marshy section lacking a defined channel in upper Reach 4 (Photo 3). Beaver activity is also responsible for a 550 m section of large ponds (approx. 1 ha) in mid-Reach 5 (Photo 4). Including this section, the proportion of pool habitat in Reach 5 represented 55% of total wetted area, essentially equal to that in Reach 4 (Table 6). Using the diagnostic procedures prescribed for Level 1 FHAP assessments (Johnston and Slaney, 1996), this abundance of pool habitat may be viewed as good for both reaches (Table 9).

The large and concentrated pond habitat in Reach 5 constitutes a separate entity, however, as it is distinctly different than the stream itself. As seen in Tables 3 and 9, inclusion of it greatly influences all characterization of the reach. For the sake of evaluating the stream itself, it is more appropriate to exclude it. This may also apply to Reach 4, excluding the more dispersed beaver impoundments dominating the pool habitat there. As shown in Table 8, when the large beaver ponds are excluded, the rating for pool abundance in Reach 5 is poor. The same may well apply to Reach 4. Pools are also lacking (poor) in Reach 6 where beaver activity is not significant.

In all of these reaches, flowing stream habitat at the time of observation was dominated by riffles (Table 6). Due to small substrate sizes (gravels, small cobbles), these riffles tended to be particularly featureless in Reaches 4 (Photo 5) and 5 (Photo 6). Boulder cover is almost non-existent in these reaches (Table 9). It is more abundant in Reach 6 (Photo 7), but even here its extent rated poor (Table 9).

The proportion of riffle habitat in Reach 6 (77% of total area) is the highest of any stream section surveyed in 1997 (Tables 3 to 5). This may be related to the extensive elimination of riparian vegetation and other forest harvesting activities along this reach (Table 4).

This also explains the particularly low (poor) abundance of LWD within Reach 6 (Table 9). It is interesting to note, however, that the cover provided by wood materials is actually higher (marginally) in Reach 6 than in Reaches 4 or 5. In both Reach 4 and Reach 5 the overall abundance of LWD is considerably greater than it is in Reach 6, and although the rating is fair in both cases (Table 9), this is actually borderline to good (> 2 pieces/bankfull channel width), especially in Reach 4.

The higher cover value of LWD in Reach 6 appears to be due to the greater stability of such elements within this reach. These frequently take the form of crosslogs (Photo 8) typical of cobble cascade-pool channel morphology (Johnston and Slaney, 1996). However, the low proportion of pool habitat, despite such structures, suggests ongoing aggradation within Reach 6, consistent with the elevated bars observed in Reaches 4 and 5 (Table 3).

Table 8. Diagnostic evaluation of fish habitat conditions in Fontaine Creek, Reaches 4 to 6, based on surveys conducted in October 1997.

Including the contribution of LWD, the overall availability of overhead cover in Reach 6 appeared to be good under the conditions observed (Table 8). This is less the case in Reaches 4 and 5, and relative to established criteria (Johnston and Slaney, op. cit.) total overhead cover may be viewed as fair in both of these reaches (Table 8). Again, though, these ratings are borderline to good (> 20%) in each case. Nonetheless, pool and glide habitats (notably the latter) often lack complexity in both of these reaches (e.g. Photo 1).

In terms of spawning habitat, the abundance of gravels is excellent throughout Reaches 4 to 6 of Fontaine Creek (Table 8). These materials are very frequently dominant within the matrix of streambed substrates. Sand and other sediments are also abundant, however, and they substantially reduce the viability of spawning at many locations. There is great abundance of gravels overall, and although constraints are widespread it may be assumed that spawning potential is more than adequate.

For resident rainbow trout, which traditionally spawn in spring and/or early summer, some side channel areas may optimize spawning potential (e.g. avoidance of high mainstem flows). These are not overly abundant anywhere within Reaches 4 to 6 of Fontaine Creek (Table 8). The highest frequency of such habitat is in Reach 5, but the greatest abundance (m side channel/m mainstem) is in Reach 6 (ibid.), where it may be viewed as marginally good. Lower proportions in Reaches 4 and 5 are likely related to beaver activity (i.e. impoundments) within the broader/flatter stream sections. Beaver dams also give rise to secondary channels in these areas (Photo 9), consistent with their effects in terms of braiding (Table 3).

Eroding banks were the most frequent disturbance indicator observed in Sovereign Creek (Tables 3 and 4), reflecting the heightened instability of this stream. Many bank failures and slumps appear to be both recent and ongoing, particularly in Reaches 1 (Photo 10) and 2 (Photo 11). At one location in Reach 2, a large section of the entire valley wall has been seriously destabilized as a result of logging disturbances coupled with steep topography (Photo 12).

Substantial failures are also present along Reach 1 of Sovereign Creek (Photo 13). This photo also shows the elevated bars that occur frequently in this reach (Table 3). An exceptionally high proportion of riffle habitat, both by length (70%) and area (72%; Table 5) again accompanied this.

Consistent with this, the proportion of pool habitat in Sovereign Creek Reach 1 is particularly low (6.6% and 5.9%, respectively; ibid.), and may certainly be viewed as poor (Table 9). Conditions are even worse in Reach 4B, where pools constitute only 2.5% of the total wetted area of the reach (Table 7). This is in marked contrast to Reach 5, where 22% of the total wetted area is pool habitat (ibid.).

Excluding consideration of the beaver impoundments in Reaches 1 and 2 of Fontaine Creek, the abundance of pool habitat in Reach 5 of Sovereign Creek is the highest in any of the surveyed streams (Tables 5 to 7). Nonetheless, even this may be viewed as poor (Table 10).

The key point is that pool abundance and frequency increase both upstream and downstream of Reach 4B. In Reach 2, pool habitat is almost as abundant (relative to wetted area) as it is in Reach 5 (Tables 9 and 10). The different channel morphology of Reach 4B is undoubtedly related to some extent to the difference between the gradient of this reach (avg. 2.5%), and all other surveyed sections of Sovereign Creek (approx. 1%; Tables 6 and 7). However, the same can hardly be said for comparisons between Reach 4A (avg. 0.8%) and Reaches 2 (avg. 1.2%), 3 (avg. 0.9%) and 5 (avg. 1.1%; ibid.).

The latter comparisons consistently indicate higher rates of occurrence and/or greater diversity of disturbances in Reach 4A than in all others excluding Reach 1 (Tables 3 and 4). The same applies to the proportion of pool habitat (7.5%; Table 7). In this regard, it is very interesting to note that this attribute is only marginally better in Reach 3 (8.5%), than in Reach 4A (Table 6).

All indications seem to suggest that Reach 4B is a major factor in the aggradation of the system, conveying effects to Reach 4A and, to a lesser extent, Reach 3. It is highly significant that both of the latter reaches are bordered principally by mature forest (Tables 6 and 7), and presumably have suffered little if any direct disturbances of their own.

There may be further implications to these findings, as well. The extent of disturbances appeared to be highest in Reach 1, as would be expected (Table 3), but as noted above indications of aggradation are most severe in Reach 4B (e.g. proportion of pools; Tables 6 and 7). This may indicate growing aggradation within this part of the system, and have major significance with respect to the potential (or even the desirability) of attempts at restoring fish habitat at the present time.

The existing complexity and diversity of rearing habitat is generally fair to poor throughout Reaches 1 to 5 of Sovereign Creek (Tables 9 and 10). Riffles and glides again tended to be rather featureless due to the predominance of gravel/cobble substrates, lack of instream LWD and disruption of riparian vegetation (Photo 14). All such conditions are somewhat superior in Reach 5 (Photo 15), but even here the complexity of rearing habitat is only fair to poor (Table 10).

As noted in the presentation of results for Fontaine Creek, two attributes specific to wood materials are included in the Level 1 FHAP diagnostic procedures. One is the percent of wood cover in pools, which provides a direct indication of the abundance and performance of wood materials, as fish habitat, within the stream. This rated fair to poor in all surveyed sections of Sovereign Creek (ibid.).

The extent of such cover is considerably higher in Reach 5 (20.3%). This value might even be viewed as good, given the established criterion of 20% (Johnston and Slaney, 1996), but the distribution of such cover tended to be clumped, and most pools investigated actually contained £ 10%. Under such circumstances, overall conditions can only be viewed as fair, even in Reach 5 (ibid.).

The second attribute addressing wood is the total abundance of LWD within the bankfull channel width. The difference is that many such elements do not extend into the wetted channel at the lower flows upon which the assessment procedures are based. Abundance is clearly good in Reach 5 of Fontaine Creek, the reverse of findings for Reach 6 (Table 8). The implication for Reach 5 of Sovereign Creek (and most other stream sections) is that the value of LWD as fish habitat may be relatively limited despite the abundance of such materials within the full channel (Photo 16).

Table 9. Diagnostic evaluation of fish habitat conditions in Sovereign Creek, Reaches 1 to 3, based on surveys conducted in October 1997.

Table 10. Diagnostic evaluation of fish habitat conditions in Sovereign Creek, Reaches 4 and 5, and Chipp Creek Reach 1, based on surveys conducted in October 1997.

The difference may be that there is little opportunity for LWD to become anchored within Sovereign Creek due to its greater instability. The displacement of such materials during elevated peak flows is a characteristic development following forest harvesting (e.g. Cederholm et.al., 1997; Slaney and Martin, 1997).

In Sovereign Creek this phenomenon is also evidenced by the substantial size and number of log jams, in Reach 1 in particular (Photo 17) and Reaches 2 and 3 as well (Table 3). These jams constitute a substantial abundance of LWD in total (fair to good; Table 9), but relatively little of it provides cover for fish at low flows (poor for all three reaches; ibid.).

To reiterate, habitat within the first five reaches of Sovereign Creek is greatly dominated by rather featureless riffles and glides (e.g. Photos 10, 12, 5, 6, 13, 14 and 15, all previously referred to). The lack of pool habitat coupled with limited cover greatly reduces the rearing potential for juvenile salmonids, including chinook, rainbow and bull trout.

There may be greater limitation of spawning potential in Sovereign Creek than Fontaine Creek Reaches 4 to 6 due to higher proportions of both cobble and sediments (primarily sand). However, gravel abundance is high (good, throughout Reaches 1 to 5), and spawning potential is likely fair (at least) throughout this stream (Tables 9 and 10).

Both logjams and beaver dams could pose problems to the upstream passage of fish in Reaches 1 and 2. This may be critical for chinook salmon (recorded for this stream) ascending the system at lower flows in late summer or fall. The same might also apply to immigrant migratory bull trout (if present), which also spawn in the fall.

Consistent with the greater size of Sovereign Creek (Tables 6 and 7), there are fewer beaver dams affecting the mainstem than in Fontaine Creek. Beaver activity appears to be confined mainly to side channels and backwaters within the Sovereign Creek system. This may be valuable in providing off-channel impoundments (potentially important rearing areas), but it may also interfere with or prevent access of adult and/or juvenile fish to such areas. It may also interfere with the delivery and/or flow of water to (and through) these areas.

A variety of such developments were observed in Reaches 1 and 2 of Sovereign Creek during the 1997 investigations. Generally speaking, off-channel habitat is surprisingly limited (fair to poor) throughout Reaches 1 to 3 (Table 9), despite the low gradient.

It is interesting that both the frequency and magnitude of off-channel habitat is considerably higher in Reaches 4 and 5. This might reflect the isolation of off-channel habitats in the lower reaches, another effect of channel aggradation (Slaney and Martin, 1997). However, although one isolated oxbow was identified for site-specific evaluation in 1997 and some isolated side channels were also observed during field investigations (Appendix 1), there is no record of these or any other such sites as disturbances (Tables 4 and 5), and their occurrence was low.

The survey of Chipp Creek Reach 1 was just over 1.5 km in length (FHAP Map # 93A-091). As noted earlier, the riparian zone of this stream section was severely disrupted during earlier logging. The whole reach is now dominated by mixed second growth at the pole-sapling stage (Table 7).

This section of Chipp Creek is very similar to Reach 6 of Fontaine Creek (discussed earlier), which has also been logged (Photo 18: Fontaine R6, Photo 19: Chipp Creek). Channel dimensions and gradient are quite similar, and both stream sections are small and relatively steep (3.2% and 4.5%, respectively) compared to other sections surveyed (Tables 5 to 7). The dominant channel morphology in each case is cobble cascade-pool (Tables 5 and 7).

What is most interesting is the similarity in fish habitat values. The proportion of pool habitat in Chipp Creek at the time of observation (10.5% of wetted area) was virtually identical to that of Fontaine Creek Reach 6. Both may be viewed as poor in terms of rearing potential (Tables 8 and 10).

The lack of pool habitat in Chipp Creek is accompanied by a great predominance of riffles (70% of area; Table 7), to much the same extent that it is in Fontaine Creek Reach 6 (77%; Table 5). Other signs of aggradation or disturbances are exceptionally low in both of these stream sections (Tables 2 and 4).

The overall availability of cover for fish is highly similar, and is marginally good in both stream sections (Tables 8 and 10). The average abundance of boulder cover in riffles is identical, but may be viewed as poor in each case (ibid.). Wood cover in pools is good in Chipp Creek and superior to that in Fontaine Creek Reach 6. The same applies to the overall abundance of LWD (ibid.). In addition to the typical cross-log elements observed in Fontaine Creek Reach 6, the base flow channel of Chipp Creek also contained an abundance of other wood elements, both stable and transient (Photo 20).

For the same reasons provided for other stream sections, spawning potential may be viewed as fair in Chipp Creek Reach 1, as it is in Fontaine Creek Reach 6 (Tables 8 and 10). Due to the presence of small falls on both of these stream sections, conditions for upstream passage may be viewed as fair (ibid.). The recorded obstructions should pose little difficulty to adult chinook and bull trout, but the same might not apply to resident rainbow trout. They would certainly prohibit immigration of chinook fry from the Sovereign Creek mainstem.

The major discrepancy between Chipp Creek Reach 1 and Fontaine Creek Reach 6 relates to the extent of off-channel habitat. In the latter stream section the abundance of such habitat relative to mainstem length (0.26 m/m mainstem) is the highest encountered in Fontaine Creek, and is judged to be good (Table 8). In Chipp Creek Reach 1 the frequency of off-channel areas (5.1/km) is nearly twice that in Fontaine Creek (2.6/km), but the total length of such habitat (0.08m/m mainstem) is poor (Table 10).

The investigation of Eskridge Creek was conducted as a short Level 1 assessment. It commenced in Reach 3, 1.2 km upstream of Reach 2 (FHAP Map # 93H-001). The channel was investigated for a distance of 500 m upstream of this location. The lowermost 250 m of the investigated distance was contained within undisturbed mature coniferous forest. The upper 250m is bounded by young (sapling stage) deciduous second growth. Clearcut logging along and including both streambanks occurred here in the late 1980’s or thereabouts.

The mean gradient of this section is approximately 4.5%, similar to that of Fontaine Creek Reach 6 (Table 5). There are other similarities between these stream sections as well. As shown in Photo 21, the channel proportions of Eskridge Creek Reach 3 are comparable to those of Fontaine Creek Reach 6. The same applies to the high proportion of boulders within the matrix of streambed materials. The predominant channel type is cobble cascade-pool in both cases.

Most importantly, LWD plays a vital role in the formation and maintenance of pool habitat in both stream sections by serving as crosslogs (Photo 8: Fontaine Creek Reach 6, previously referred to), and deflectors (Photo 21: Eskridge Reach 3, referred to above). These functions are most evident in the undisturbed section of Eskridge Creek, where more frequent occurrence of such elements resulted in a higher frequency and abundance of pool habitat.

There are no indications of severe channel disturbances in the section of Eskridge Creek Reach 3 subjected to clearcut logging (e.g. Tables 2 to 4). The abundance of LWD is distinctly lower, though, as is the frequency and extent of pool habitat. As in Fontaine Creek Reach 6, macrohabitat in the disturbed section of Eskridge Creek is greatly dominated by riffle.

The above analogies between Eskridge and Fontaine creeks apply to Chipp Creek Reach 1. It is also characterized by cobble cascade-pool morphology, with the occurrence of pools largely controlled by the presence of stable LWD (Photo 19, previously referred to). While such elements are more abundant here than they are in Fontaine Creek Reach 6 (discussed previously), macrohabitat is still greatly dominated by riffle (Table 7).

The findings in Eskridge Creek emphasize the profound importance of stable LWD to the formation and maintenance of pool habitat in such streams. They also reflect the loss of it and associated habitat values as a result of harvesting the riparian areas along streams.

Such harvesting may initially result in substantial introduction of woody debris to affected channels. This is seldom stable however, and with the high peak flows typically associated with clearcut logging it is ultimately displaced. The same peak flows also dislodge and displace formerly stable LWD (Cederholm et. al., 1997).

What is worse is that the elimination of mature coniferous trees along stream banks precludes the natural replacement of adequately large LWD for at least several decades (ibid.). Implications in terms of fish habitat are obvious, and were well evidenced in the 1997 surveys. While deciduous regeneration is relatively fast growing, it tends not to reach sizes comparable to the mature conifers that typically provide the stable LWD in B.C. systems. Furthermore, deciduous species are less resistant to rot, and their instream durability is limited (ibid.).

Even though the abundance of functional debris has been diminished in Fontaine Creek Reach 6 and Chipp Creek Reach 1, these stream sections still provided the best rearing habitat for fish of all stream sections subjected to Level 1 assessment (Tables 8 to 10). This is particularly true for Chipp Creek, where there is greater retention of LWD. The investigation of Eskridge Creek Reach 3 is further confirmation of the great significance of these materials in such streams, and of stream sections investigated in 1997 it undoubtedly provides the best natural template for habitat rehabilitation proposals.

Lists of all site-specific assessments conducted during the October 1997 investigations are in Tables 11 (Victoria Creek drainage), 12 (Fontaine Creek drainage) and 13 (Sovereign Creek drainage). Locations, concerns to be addressed and general conclusions are also provided for each site.

Investigations in the Victoria Creek drainage principally addressed potential obstructions to fish passage within Reaches 1 to 3 of the mainstem. Log/debris jams, occasionally with associated beaver dams, were inspected at nine locations (Table 11). Logging debris was rarely present in any of the jams assessed. At the time of inspection, none of these obstructions was judged to be a barrier to upstream passage of fish. In most cases the log/debris accumulations are permeable (e.g. Photo 22: Reach 1), and all beaver dams could be bypassed, if not ascended.

At some locations there are higher probabilities that in the future more severe blockages might develop with further accumulation of materials (Table 11). In most such cases, however, the shallow topography of riparian areas would result in the formation of bypass channels (Photo 23: Reach 3). Unfortunately, the potential to form total barriers appeared to be highest for the two jams inspected in Reach 1 (Photo 24). Implications with respect to chinook salmon are obvious, and the same might well apply to migratory bull trout if they are present within this drainage.

Fish passage concerns were also addressed at a culverted road crossing of a small unnamed tributary to Reach 2 of the Victoria Creek mainstem, 2.2 km upstream of Reach 1 (site V-3; Table 11). This tributary falls within Weldwood of Canada Ltd.’s TFL. Electrofishing below this culvert produced both chinook and rainbow trout. No fish of any species were captured above the culvert. Habitat upstream may be viewed as marginal at best. For the first 40 m, gradient exceeds 20%. It then levels out for approximately 300 m, but flows disappear periodically within the bed materials of this section. It seems likely that this portion of the stream (at least) would freeze entirely during winter.

In the course of the culvert inspection, field personnel observed severe impacts of forest harvesting on the section of stream below the culvert. The first 200 m downstream is bordered by a clearcut, and is severely choked with silt and unstable debris. The debris is so extensive and movement on foot is so difficult that a Level 1 FHAP would not be possible if it were desired.

The remaining 200 m of stream flows through mature coniferous forest, and includes the stable LWD elements discussed previously. This section will likely deteriorate without intervention, as it is clearly suffering intrusions of unstable debris and silt from the cutblock upstream. Attention to this should outweigh any further consideration of the culvert crossing itself.

Table 11. Site-specific investigations conducted within the Victoria Creek drainage; October 1997.

Table 12. Site-specific investigations conducted within the Fontaine Creek drainage; October 1997.

Table 13. Site-specific investigations conducted within the Sovereign Creek drainage; October 1997.

The final site-specific investigation within the Victoria Creek sub-basin involved the assessment of the Robertson Lake watershed (site V-11; Table 11). It was identified in the 1997 Overview FHAP as a possible source of the woody debris forming jams in the Victoria Creek mainstem. This system drains to Victoria Creek near the top of Reach 3 (FHAP Map # 93B-090). It has been subjected to substantial forest harvesting in the 1970’s and 1980’s.

Inspection of upper reaches in this system indicated apparently good recovery of streams from recent logging. Although riparian vegetation is sparse, the occurrence of stable LWD remains adequate, and a good diversity of riffle-pool habitat has been maintained. Problems do exist further downstream in the lowermost 1 km of stream length, immediately upstream of Victoria Creek. This section does contain a considerable amount of unstable woody debris. There are several barrier jams, and in addition to logging-related debris there is also frequent Riparian Reserve Zone blowdown within and across the channel. Although these areas are undoubtedly contributing some organic debris into Victoria Creek, they are not significantly affecting jams downstream.

On the strength of the 1997 Overview FHAP for the Victoria Creek sub-basin, one other site was scheduled for Level 1 inspection. This is an agricultural/ranch homestead site on Reach 6, where disruption of streambanks is evidenced in aerial photography (Carmanah, 1997). Unfortunately, the requisite permission for access could not be obtained for this site.

The first site-specific investigation on Fontaine Creek (site F-1; Table 12) was at a channelized section of the mainstem immediately downstream of the bridge crossing at the top of Reach 1 (FHAP Map # 93A-092). This section is approximately 100 m in length, and is dominated by swift-flowing riffle habitat when observed in October 1997 (Photo 25).

Rather sparse riparian vegetation is at the pole-sapling stage along both banks. It is dominated by deciduous species, and provides overhead cover to approximately 10% of the total wetted area of the channel. Other cover is severely lacking, and is principally limited to shallow cutbanks. The banks themselves are bermed (approx. 1 m) along both sides of the channel (Photo 26).

This site was investigated for potential installations of fish cover elements to provide refuge from laminar flows and create pool habitat, if possible. This section of the stream is bordered by flat low-lying land, which might provide the opportunity for side channel and/or backwater development. Several settling ponds associated with past placer mining are already present, and they could be integrated with such an undertaking (Photo 27). In addition to excellent road access and flat terrain, construction (involving machinery) would be facilitated by the lack of vegetation beyond the narrow (5 m to 10 m) riparian buffer on both sides of the channel (Photo 28).

The second location inspected on Fontaine Creek (site F-2; Table 12) was an abandoned placer mining site on the west side of the mainstem, 400 m downstream of the top of Reach 2 (FHAP Map # 93A-092). Three large settling ponds are present at this location. The chief interest and concern is the substantial extent of disturbed and eroding banks resulting from the mining operations (Photo 29).

Although these banks are set 30 to 50 m back from the channel of Fontaine Creek, there is evidence of surface runoff to the stream during high flow periods. Such runoff may result in substantial inputs of silts and other fine sediments. The opposite (east) bank of Fontaine Creek is eroding at this location, another evident source of fine sediments.

A second placer mining development located 250 m upstream of site F-2 was also identified for inspection as a possible source of concentrated sediment inputs (site F-3; FHAP Map # 93A-092). No significant disturbance could be found at this location. The only evidence of past mining activity was an old access road and some discarded plastic piping.

The next site inspected (F-4) in the Fontaine Creek drainage was on Horan Creek, the mouth of which designates the break between Reaches 4 and 5 of the Fontaine Creek mainstem (FHAP Map # 93A-092). In the 1997 Overview FHAP of the Fontaine Creek drainage, substantial impacts of logging were noted in upper Horan Creek and recommended for further investigation (Carmanah, 1997).

This was conducted in upper Reach 2, 3 km upstream of the Fontaine Creek mainstem (site F-4; FHAP Map # 93A-092). At this location nearly all mature riparian forest has been eliminated, and the channel itself is rather simplistic (Photo 30). Width is in the order of 3 m, and gravels (50%) and cobbles (45%) are the dominant bed materials. Fines content is relatively low (5%).

There is a limited abundance of LWD, and it is only occasionally functional. In all respects, there appears to be a great deficiency in cover. Rainbow trout are reported to be present within this part of the system (Carmanah, 1997). Existing conditions would not favour the maintenance of fish production, given the lack of refuge during flood events, the exposure to predators and the possibility of excessive temperatures during summer months (Slaney and Martin, 1997). The Riparian Assessment Procedure (Carmanah, 1997) identified this area as being sufficiently restocked, and therefore the riparian function will remain compromised until the vegetation matures.

It is possible to develop complex side channel habitat at site F-4. As at site F-1 in Reach 1 of Fontaine Creek, the flat terrain at site F-4 on Horan Creek can accommodate such development (Photo 30; referred to above). However, access conditions are somewhat less favourable.

The last site inspected within the Fontaine Creek drainage was a culvert crossing on an unnamed tributary which designates the break between Reaches 5 and 6 of the Fontaine Creek mainstem (site F-5; FHAP Map # 93A-092). The culvert is located approximately 300 m upstream of Fontaine Creek. It had a 0.6 m vertical drop at its downstream end under the flow conditions observed (Photo 31).

Sampling by electrofishing produced juvenile rainbow trout both downstream and upstream of this culvert. While this does not necessarily confirm that it is passable for fish, it is consistent with all other impressions that it is, even for small resident adults.

The first site-specific inspection on Sovereign Creek was an inactive channel section believed to be an isolated oxbow. This is located on Reach 2 of the Sovereign Creek mainstem, 2.6 km upstream of Reach 1 (site S-1; FHAP Map # 93B-100). The channel is 50 to 75 m in length, and may provide some opportunity for development.

Indications would suggest, however, that it serves as a flood channel (at least) during higher flows. Standing water was present in depressions at the time of observation. The magnitude of flows and the role of this channel during flood events should be ascertained before any modification of it is proposed.

The second site inspection on Sovereign Creek was a side channel adjacent to Reach 4 of the mainstem, 500 m downstream of Reach 5 (site S-2; FHAP Map # 93A-091). This channel presently provides excellent habitat at low flows (Photo 32), but an additional 150 m might be developed with augmentation of water supply from the mainstem. The addition of cover would not be required, as abundant LWD is already present.

A long since abandoned mill site was the third to be inspected in this drainage. It is located on Reach 5 of Sovereign Creek, 400 m upstream of Reach 4 (site S-3; FHAP Map # 93A-091). This location was suspected to be a source of concentrated sediment inputs to the mainstem (Carmanah, 1997).

No such indications were observed at the site. However an active placer mining site was located directly downstream just below the bridge crossing. There are extensive areas of disturbed and unvegetated slopes at this location. This site should be stabilized and revegetated once mining operations cease.

The two remaining site-specific inspections on Sovereign Creek also addressed placer mining developments. The first was on Eskridge Creek, close to its mouth (site S-4; FHAP Map # 93A-091). This operation appears to have been abandoned for a lengthy period of time, and conditions are relatively stable. Much of the disturbed area is now revegetated with grasses, shrubs and coniferous saplings. There is some erosion of the streambanks, but ongoing impacts are judged to be low (Photo 33). The streambanks are low for approximately 100 m, and siltation may occur with overtopping during freshets. Even in this event, potential impacts are judged to be of a low priority in terms of rehabilitation requirements.

Conditions at the last site are another matter altogether. This site was located on an unnamed stream that is tributary to Reach 6 of Sovereign Creek, approximately 1 km upstream of Eskridge Creek (site S-5; FHAP Map # 93A-091). Once again, the site was located near the mouth of the stream, and it is undoubtedly a significant source of sediment inputs to the Sovereign Creek mainstem.

The stream channel flows through the earlier mine site, and may have been diverted. It is unstable and actively eroding over a length of approximately 200 m. The banks are largely composed of tailings, and are not resistant to erosion. Conditions are further aggravated by moderately steep gradient (Photo 34), and in the absence of intervention it is unlikely the site will stabilize within the foreseeable future.

It is apparent that the most extensive degradation of habitat evidenced in the 1997 surveys within the Cottonwood River drainage is in Sovereign Creek. This is consistent with the impressions of 1997 Overview FHAP assessment (Carmanah, 1997). For this reason, Sovereign Creek will be the focus of attention in following consideration of appropriate rehabilitation objectives and techniques. Due to similarities of the basic geomorphology in all cases, these comments are also relevant to all other streams investigated in 1997. Varying conditions and emphases will be addressed as appropriate.

In the introductory sections of this report it is emphasized that all final identification of specific techniques, site selection and risk assessment of instream rehabilitation activities must be conducted by a fully qualified fluvial morphologist or hydraulic engineer (Allan and Lowe, 1997; Cederholm et.al., 1997; Slaney and Martin, 1997). This applies most emphatically to the Sovereign Creek drainage due to its very apparent instability.

In such cases, instream structures may cause more harm than good. As a rule of thumb, unstable areas should be avoided when considering any such undertakings (Cederholm et.al., op. cit.). The most urgent need for Sovereign Creek is restoration of geomorphic stability, as evidenced by the extensive erosion and failure of banks, particularly in Reaches 1 and 2 (Table 3). Until the system has stabilized sufficiently, attempts at instream rehabilitation may be fruitless or even counter-productive (ibid.). Structures may lack effect or durability, and could even lead to further destabilization of the channel.

Also emphasized earlier, the restoration of stability in logged watersheds can only be achieved through rehabilitation of the watershed as a whole. This commences with all relevant upslope areas (e.g. Slaney and Martin, op. cit.). In the case of severely destabilized systems, such work should precede any instream activities so as to ensure that the instream work is both beneficial (as opposed to detrimental) and successful (ibid.). Again, all following comments and recommendations assume that all requisite upslope and riparian rehabilitation will be undertaken expediently for all streams addressed.

Especially in the case of Sovereign Creek, it is assumed that such rehabilitation of upslope and riparian disturbances will be pursued as the top priority for the system. This being the case, there may then be the opportunity to accelerate recovery of the stream channel through concerted efforts at bank stabilization. In Reaches 1 and 2 alone, severe bank disturbances were recorded at over 30 locations (Table 3). To varying degrees, there may be opportunities to stabilize these areas employing various techniques suggested under the WRP guidelines (WRP Tech. Circ. No. 9). Wherever the need for such intervention exists, it should be pursued as a top rehabilitation priority, second only to upslope and riparian rehabilitation.

For conditions in Sovereign Creek, the most appropriate techniques would likely be brush mattresses, vegetated geogrids and live cribwalls (Babakaiff et. al., 1997). For particularly steep banks, and/or those that cannot be sloped back with machinery (feasibly or practically), the latter two techniques may prove to be most applicable and successful (ibid.).

Once greater channel stability has been achieved, there should be more opportunities to install instream structures to improve the abundance of cover and hydraulic diversity for fish with less risk of failure and detrimental effects. The creation of additional pool area and the augmentation of cover are the greatest fish habitat rehabilitation needs throughout Reaches 1 to 6 of Sovereign Creek (Tables 3 and 4). Such habitat is of exceptional value to all salmonid species present within the system (chinook salmon, rainbow trout and bull trout).

Excluding the role of beaver activity, which will be addressed below, LWD is certainly the most important factor in the formation of pools and provision of cover in all of the stream sections surveyed in 1997. This is the norm for streams of this type, where channels are dominated by smaller bed materials (Hogan and Ward, 1997). The predominant role of LWD was clearly evidenced in all of the less disturbed stream sections that were investigated (e.g. Eskridge Creek, Fontaine Creek Reach 6, Chipp Creek).

Using nature as the template, consistent with FHAP procedures (Johnston and Slaney, 1996), LWD additions would seem to be the most appropriate method of increasing both pool habitat and cover in virtually all of these streams. In Fontaine Creek, where bank and channel stability is superior to that of Sovereign Creek (Table 2), there appear to be abundant opportunities at the present time for the development of pool habitat (abundance and frequency) and cover through LWD additions. Such rehabilitation is the priority for fish habitat in Fontaine Creek (Table 8), Sovereign Creek and elsewhere (Tables 9 and 10). The pool habitats caused by beavers in Reach 4 and lower Reach 5 of Fontaine Creek (Table 8) are a special circumstance, and will be addressed below.

It must be emphasized that in unstable channel environments the introduction of LWD may add to bank erosion and instability (Cederholm et. al., 1997). Even in Fontaine Creek, such installations should only be considered for the most stable areas. In Sovereign Creek the extent of instability directly limits acceptable opportunities under present conditions, even though widespread installations might ultimately be considered for this stream.

On the basis of the 1997 field data and photographic record, a great number of locations for LWD installations have been tentatively identified in Sovereign Creek and other stream sections. These are detailed in Appendix 1. Particular attention has been focused on the identification of long riffle and glide sections, where the interspersion of pool habitat would be most desirable. This process of identification was based principally on biological requirements and objectives, with consideration of obvious geomorphic implications. However, many of these sites may prove to be unsuitable or unacceptable when subjected to the requisite engineering evaluation.

Nonetheless, under present conditions there may be a limited number of sites where LWD materials could confidently be installed within the Sovereign Creek mainstem. Where such opportunities do exist, they would provide valuable experimentation for future work in this stream. The greatest opportunities appear to exist in Reaches 3 and 5, where conditions are more stable (Table 6).

In terms of priorities between systems, LWD additions in Fontaine Creek would be given precedence over those in Sovereign Creek, especially from the standpoint of channel stability and associated likelihood of success (Cederholm et.al., op cit.). From the fisheries perspective, however, higher priority might be given to Sovereign Creek due to the presence of chinook salmon (and possibly bull trout) in addition to the rainbow trout present in Fontaine Creek.

As noted earlier, the extensive beaver activity in Reach 4 and lower Reach 5 of Fontaine Creek presents some interesting implications specific to this system. In a stream supporting anadromous stocks, the immediate response to beaver dams and impoundments may be to eliminate them so as to minimize risks to fish passage. In Fontaine Creek, which is reported to contain only resident rainbow trout, the removal of beaver dams could be severely counter-productive. As emphasized in the presentation of results, these structures provide an excellent abundance of pool habitat (Table 8), which may be particularly productive for trout (Finnigan and Marshall, 1997).

Certainly, these dams and impoundments might preclude internal migrations of small resident fish, but on the other hand, they may not; upstream migrations of rainbow trout (spring spawners) would likely occur during higher water when dams might be easier to ascend and/or bypass.

In essence, removal of the dams on Fontaine Creek would represent the opposite of the foremost habitat rehabilitation objective (the creation of greater amounts of pool habitat) for this and all other streams investigated in the 1997 assessments. It is acknowledged that beaver impoundments are not the equivalent of the small fluvial pools that are the ultimate objective of stream rehabilitation, and they certainly do not provide the frequency of pool habitat that is desired (Table 8). But even if only as temporary compensation, they offer great potential for the augmentation of complex rearing habitat through additions of LWD cover. Great amounts of such cover could be installed in the beaver impoundments within Reach 4 and lower Reach 5 of Fontaine Creek, and this could be accomplished by helicopter with relative ease.

This is not suggested as an alternative to the development of other pool habitats in this stream. In fact, a variety of possible opportunities similar to those in Sovereign Creek have been tentatively identified in upper Reach 5 and Reach 6 of Fontaine Creek, as well (Appendix 1). These should be pursued as the higher priority in terms of stream rehabilitation, per se. However, if there is a concern about safeguarding or maximizing the production of resident fish within this stream, the beaver impoundments would likely offer the greatest potential.

For LWD installations in flowing stream sections, various techniques are fully addressed by Cederholm et.al. (1997) and Slaney et.al. (1997) in the WRP guidelines (WRP Tech. Circ. No. 9). Further discussions and guidelines are also provided by Anonymous (1986; 1989). Different techniques will be required, or will be most appropriate, in different locations. To the extent possible, it would probably be most beneficial to attempt replication of the crosslog elements so closely associated with pool formation and maintenance in less disturbed sections. This might be accomplished with log weirs (Allan and Lowe, 1997), but stream width and/or gradient may be excessive for this technique at many locations (ibid.). The same might also apply to all other LWD techniques, as well (Cederholm et.al., op. cit.).

Again, a qualified and experienced fluvial morphologist or hydraulic engineer must address all such issues on a site-specific basis (ibid.). The introduction of boulder groupings may provide an alternative to LWD in some cases (Ward, 1997), but this technique is not recommended for aggrading channels or streams (or stream sections) where such habitat does not occur naturally (i.e. is not part of the natural template; Hogan and Ward, 1997). Aside from beaver activity, LWD is the principal determinant of pool abundance and frequency in the streams surveyed within the Cottonwood River drainage in 1997. Consideration of boulder additions is only advised for a limited section of lowermost Sovereign Creek Reach 1, where such materials do occur naturally (Appendix 1).

With respect to LWD structures other than weirs, the availability of adequate streambank anchors may be another constraint, particularly in view of the extensive removal of mature coniferous trees along many stream sections. In such cases it might be advantageous to consider the use of reef-type LWD installations (Slaney et.al., 1997), as opposed to more traditional methods. These might be particularly suitable for streams addressed here, as they are known to be utilized by spawning chinook in addition to both chinook and trout juveniles (ibid.).

On the other hand, such structures may be prone to infilling with bedload, and are principally recommended for relatively stable streams with advanced upslope recovery (ibid.). Again, this may limit their utility in Sovereign Creek, in particular. Infilling may be avoided to some extent by installing the reefs in areas with adequate hydraulic energy. The technology is new, and there is much to be learned about the potential uses for these structures and their associated values. Accordingly, some level of experimentation with them might be valuable, even in Sovereign Creek.

In cases like Sovereign Creek, where ongoing instability diminishes the potential for (and/or advisability of) structure installations in mainstem environments, development of off-channel habitats may offer a temporary or permanent alternative (Lister and Finnigan, 1997). A substantial number of such opportunities was tentatively identified during the 1997 field surveys of Sovereign Creek, and the other streams as well (Appendix 1).

Side channels and backwaters can be extremely important for salmonids, especially coho salmon (ibid.). However, such habitat tends not to be used as heavily by chinook, rainbow trout, or bull trout (ibid.). Nonetheless, it may still provide critical refuge during flood events or during winter. While this may be of limited relevance to chinook, it may be very important to rainbow and bull trout. Accordingly, development of off-channel habitat may be of value, especially in Reaches 1 and 2 of Sovereign Creek, where mainstem activities would be constrained by channel instability.

Lister and Finnigan (op. cit.) emphasize that attempts to develop off-channel habitat are subject to a variety of constraints and conditions. Once again, areas that lie within active flood plains or are subject to other instability should be avoided. The slope should be at least 0.5%, and upstream mainstem intakes should be avoided.

Alcove ponds, analogous to the settling ponds on Fontaine Creek (for example), may be productive, but once again they are most suited to hydraulically stable environments where they are less likely to be cut off by bedload movement (ibid.). The same applies to any other side channel or backwater habitat that might be created, reactivated or otherwise modified in unstable systems.

In cases where such habitats have been alienated by roads or other such developments, they should certainly be reconnected, as they generally offer good stability (ibid.). One such case was identified in Reach 2 of Sovereign Creek (Appendix 1). Attention might also be paid to side channels blocked by debris jams, especially if the jams can be removed with reasonable ease and expense. In keeping with earlier discussions of beaver impoundments on Fontaine Creek, any disruption of beaver dams could be detrimental. Furthermore, any attempts to eliminate the effects of beaver activity are likely to be short-lived (Finnigan and Marshall, 1997).

Developing new off-channel habitat may be advantageous in areas unsuitable for mainstem improvements. Again, this might apply to various stream sections addressed in the 1997 investigations; some particularly promising opportunities were identified. In the name of stream rehabilitation, however, and with consideration of the species present, it would seem far more appropriate to concentrate efforts in mainstem environments to the fullest extent possible for all streams addressed in this study.

Consistent with preceding discussions, general recommendations for the rehabilitation of all stream sections subjected to Level 1 FHAP surveys in 1997 are provided in Table 14. As noted in the preceding section, a preliminary list of specific sites and opportunities for all of these sections is provided in Appendix 1. However, it is essential that all of these suggestions be further evaluated through Level 2 FHAP assessments (Fontaine Creek, Sovereign Creek, and Chipp Creek). These should specifically include geomorphic and engineering expertise to confirm (or not) the preliminary recommendations provided here and to address all specific installation sites and considerations.

Recommendations relative to all site-specific investigations conducted in 1997 are provided in Table 15. The only assessment not included in these Tables is that of Eskridge Creek Reach 3. In view of the complex and relatively undisturbed habitat below the section subjected to clearcut logging, a full scale Level 1 FHAP seems warranted for this stream. Alternatively, a Level 2 assessment might be conducted in concert with those identified above in order to expedite rehabilitation of the disturbed stream section.

In terms of biophysical urgency, priorities for various undertakings have been identified to the fullest extent possible in Tables 14 and 15. The Sovereign Creek system has definitely been the most severely disturbed and warrants the greatest concern, especially considering the presence of chinook and bull trout. However, as repeatedly emphasized, the extent to which these concerns can be reflected in operational priorities (and undertakings) must await the requisite morphometric and engineering expertise and evaluation.

With such input, it may be concluded that little instream work of any kind can be considered for Sovereign Creek at present. It may be necessary to await stabilization following rehabilitation of upslope areas. If this is the case, all available funding and effort for this stream might best be channelled to the latter priorities. Alternatively, such funds and efforts might be reallocated to Fontaine Creek, where the possibility of successful instream rehabilitation seems much higher at present.

Table 14. General fish habitat restoration recommendations following Level 1 FHAP assessments in Fontaine, Sovereign and Chipp creeks in October 1997.

Table 15. Recommendations for site-specific assessments conducted within the Victoria, Fontaine and Sovereign creek drainages in October 1997.