Loading...
HomeMy WebLinkAbout1996 Freshwater Habitat Conditions Affecting Coho Salmonr- � 7 Freshwater Habitat Conditions Affecting Strait of Juan de Fuca Populations of Coho Salmon (Oncorhynchus kisutch) Report to the Pacific Fisheries Management Council June 1996 Michael L. McHenry Fisheries Habitat Bioloaist Lower Elwha Klallam Tribe �(LEKT) 2997 Lower Elwha Road Port Angeles, Washington 98363 INTRODUCTION The Strait of Juan de Fuca (SJF) region encompasses an area of some 1,500 mil along the North Olympic Peninsula (Figure 1). The region includes a diversity of watersheds that support anadromous salmon, ranging in size from the 335 mi' Elwha River, to several watersheds less than 10 mil. The majority of these systems support sympatric populations of coho, chum salmon, and winter -run steelhead. The two largest rivers, Elwha and Dungeness, are heavily impacted by hydro - modifications, and are managed primarily for hatchery production of coho. The remainder of SJF streams are currently managed for wild coho production'. Coho production has been chronically low for the region as a whole. A wild escapement goal of 11,900 fish was established by fishery managers in 1983, and has only reached 700 of the goal in one year. As a result, Strait of Juan de Fuca (SJF) coho Populations have been identified by the Pacific Fisheries Management Council (PFMC) as "chronically under escapement goals ". The Council is concerned about the effect of ocean fisheries, habitat conditions and hatchery practices on such stocks. As part Of its biennial review process, the PFMC has ordered a comprehensive status review of SJF coho. This report summarizes the pertinent habitat issues that affect SJF coho salmon during their freshwater life history. The report is divided into three sections: 1) A introduction to the watersheds and description of the characteristics that affect natural production, 2) A summary of coho salmon life history and comparison with actual habitat conditions available in the SJF, and 3) a life history model developed from two SJF watersheds to quantify production losses. Almost every stream in the SJF has a history of our-planting non - native stocks (WDFW et al. 1994). However, because of low survival rates, the suspected effects of off likely low. - station releases is 1 SJF WATERSHED CONDITIONS Geology The geologic history of the Olympic Peninsula is complex and involves rock formations derived from the ocean floor. The collision of continental and oceanic plates provided the necessary forces for the uplifting of marine sedimentary (peripheral) and basalt (core) deposits that formed the Olympic Mountains (Tabor & Cady 1975) . The Olympics were subsequently modified by both continental and alpine glaciation. The most recent glaciation began during the Pleistocene era, when at least 6 separate lobes of the Cordilleran ice sheet invaded Puget Sound (Tabor 1987). This event', which shaped many present -day features of the region including Puget Sound, Hood Canal and the Strait of Juan de Fuca ended approximately 12,000 ybp2. The period immediately following deglaciation was likely unfavorable for salmonids, as climatic and watershed conditions were highly unstable (Benda et al. 1993). Around 4500 ybp, climatic conditions changed, becoming cooler and wetter, allowing the proliferation of coniferous forests, which stabilized watershed conditions, and ultimately salmon populations (Chatters et al. 1995). Hydrology The most conspicuous hydrologic feature of the Olympic Peninsula is the radial drainage pattern of its 11 major river systems that originate off the core Olympic Mountains. With elevations to 8000', the high Olympics directly intercept Pacific storms, and receive more rainfall than any other place in conterminous United States. Of the large river systems, only the Elwha and Dungeness drain directly to the SJF. Both rivers have been influenced by alpine glaciation, and receive significant snowpack at higher elevations. As a result, bimodal discharae peaks (one in winter associated with precipitation events, and another in spring /summer asscciated with snowmelt) are obser-red n 'years before present 2 PamM;, y State Hoko River of S7RA1T OF Washington _ Pysht River JUAN DE FUC1 1 Call B,,, - Ba L.ocaticn Map - Lyre River flwha River Ozette Lake r ¢ Dan encss River - - - _ � —� Pat Atteeles �� l •' • _ _ ( • `-� Pon To«nsend i Lake Ctescertt Pfil.Y7 _ -- .. •rll� OCEAN - _ 'l o Quilcene /I Legend _ :.., { 1133 j I • Cities / Tomru _ - r� 2�•'' Rims / So =eazTls Glaciers National Pant oa � Glaciers / Snowhelds - 1 4.5 0 4.5 Miles ° TIVS<X=scales rmncioonibioi:= iioti.000 Sde i 'M.1 0 14ujoa,orc WayV t2um 4abv _ Quu>attlt Lakes t LakeCttsfmntti }} �f Figure 2. Strait of Juan de Fuca Region, Olympic Peninsula, Washington. 3 the hydrograph. Flow patterns in combination with watershed size, and habitat diversity have allowed for a diversity of salmon species (pink, chum, coho, spring Chinook, summer steelhead) not observed in the smaller SJF watersheds. Hydrologic characteristics of the remaining SJF watersheds are influenced by their position in relation to the rainshadow of the Olympic Mountains. Watersheds located on the leeward side of the Olympics may have extremely low annual precipitation levels. For example, totals as low 15" \year are recorded in the Dungeness Valley, near Sequim. Water production per unit area in this portion of the SJF is more similar to that of southern Oregon or Northern California (Lichatowich 1993a). These conditions present a higher risk of local extinction in the SJF, similar to those seen for salmon stocks at the geographical extremes of their natural distribution. Precipitation increases dramatically along a westerly gradient in the SJF and annual precipitation averages over 100" \year near Cape Flattery. Streamflow patterns in this portion of the SJF are similar to those found on the non - glaciated streams of the Olympic coast. Peak discharges during winter freshets. The western SJF is located within one of the highest runoff yield zones in western Washington (Naiman et al. 1992). Peak flows recorded for the Hoko River are 2 -3 times those of the Dungeness River, a watershed that is 4 times larger than the Hoko. Hydrologic stress on bioioaical systems are extremely high in the SJF and multiple flood events occur annually. Summer flows tend to be naturally low, as 1itt?e snow accumulation occurs in these low elevation systems. vegetation Rainfall patterns, as affected by the Olt,,npic rainshadowa, have heavily influenced vegetation patterns in the SJF. Clima:: vegetation communities are general!% dominated by con - serous forests (Franklin & Dyrness 1984). Biomass of organic mazer-J.-.1 (both standing and dead) may be enormous in unmanaged forests. T:: lower rainfall areas of the SJF, mature forests are dominared b- 4 Douglas fir (Pseudotsuga menziesii) and western Hemlock (Tsuga heterophylla). In the western SJF, beginning around Deep Creek, vegetation becomes transitional, with proportionally greater amounts of Sitka spruce (Picea stichensis) and western red -Cedar (Thuja plicata). Historically, large openings supporting prairie vegetation communities, were maintained through the use of fire by native Americans (Gorsline 1993). These areas have been largely destroyed by agriculture and development. Disturbances to native forest communities (logging, fire) have resulted in large areas of the riparian landscape being dominated by monotvpic stands of red alder (Alnus rubra). Land Uses /An thropogenic Impacts The SJF region includes a variety of land uses that have affected coho salmon production. Human activities in riparian zones, floodplains, and cumulative effects at the watershed level in the Pacific Northwest, have led to the simplification and fragmentation of salmon habitat (Seddell & Everest 1991; Frissell 1993) . Such activities may affect a given habitats productivity or destroy its connectivity to other unaffected habitats within a watershed. Although nearly one million acres of the Olympic Peninsula is federally protected by Olympic National Park (ONP), ONP currently affords relatively little protection for SJF salmon populations'. The Park is centered around the mountainous interior of the Olympics and the upper river valleys of its major rivers. During ONP' s creation, the majority of low elevation, old- growth timber was excluded from its boundaries. As a result, almost all of the smaller SJF watersheds and much of the most productive low elevation fish habitats have been subjected to various land uses. 3A significant portion of Morse Creek above the limits of anadromous salmon is within Olympic National Park. Onlv the extreme headwaters of the Dungeness River are located within ONP. Over 80% of the E1wha River watershed is located within the ONP, however, mainstem dams prevent access at river mile 4.9 and 12.0. 5 The original old- growth stands of timber have been removed and are managed by large industrial timber owners as forest plantations. Logging In the Pacific Northwest, logging practices have historically been conducted without regards to fisheries resources. Logging began in the middle part of the nineteenth century in the eastern SJF and in the early portion of the twentieth century in the western SJF. Initial entries were associated primarily with waterborne transportation routes; shoreline and river valleys were logged first. To facilitate log transport, stream channels were cleaned of woody debris. Railroad logging was introduced to the Peninsula around 1915. Railroad spurs were usually constructed through the larger river valleys and their larger tributaries. By the 1940's, truck logging became the dominant technique. Extensive road networks were constructed in steeper, upper watershed areas, allowing access to previously inaccessible old - growth forests. Road building techniques were primitive, and these practices left a legacy of accelerated landslides. For example, in Deep Creek a total of 134 landslides were documented between 1971- 1991 (Shaw 1995). At least 335 landslides were inventoried from the available air photo record in the Hoko River (Pentec 1995). By the 1970, s, the conversion of the regions original old - growth forests to tree farms was complete. The vast majority of the regions forests are now dominated by young age- classes of trees. Road densities are often high (2 -5 mi;mi2), and logging rotations typically are between 50 -65 years. Agriculture Agricultural development, though less widespread thar forestry, has affected low - gradient habitats of the SJF These impacts include dredging, diking, and water removal. Streams with the most severe impacts include the Clallam, Hoko and Dungeness rivers, and Salt, Salmon, Snow and Chimacum creeks. For example, the lower 2.0 miles of the Little Hoko River ha-:e been channelized and cleaned of woody debris. Water withdrawals also impact several SJF rivers. Over 400 miles of irrigation canals have been built in the Dungeness valley. Surface water withdrawals on the Dungeness River currently average about 600 of the rivers natural stream flow. A total of 579 cfs could be legally removed from the Dungeness River, even though the summer low flows average around 200 cfs. Channelization /Urbanization Population growth has also impacted coho salmon in the SJF. Channelization and flood control efforts to accommodate flood plain developments have severely altered the Dungeness River, Morse, Salmon, Snow and Chimacum creeks. The majority of these activities are historic, dating to the early part of the twentieth century (Bahls & Rubin 1996) . Channelization isolates floodplain habitats from mainstems, preventing coho salmon from reaching off - channel rearing areas critical to over- winter survival. Channelization also increases stream energy, resulting in loss of pool habitat and coarsening of stream substrate. Channelization effects have been exacerbated by the systematic removal of woody debris and loss of riparian forests across the region. Dick Goin, a long -time resident, offered a poignant description of the effects of channelization on the Dungeness River (Lichatowich 1993b): "One of the reasons I quit fishing the Dungeness River was the bulldozers. The 1960s was the age of the bulldozers. Seems like we would have a nice hole for fishing and then is would be ground out by a bulldozer. When I came back 10 years later the destruction was almost total. All the old structures were gone. I think it was the mid -70s before there was an effort to stop them. Incredible damage was done by then." The combination of channelization and water withdrawals have devastated Dungeness River coho populations. In terms of potencial production, the Dungeness and its tributaries represent 1S.6a of the total SJF smolt vield. Urbanization, particularly around Port Angeles, has impacted six small independent tributaries to the SJF: Morse, Lees, Ennis, 7 Peabody, Valley, Tumwater, and Dry creeks. Impacts to habitat in these streams are so severe that coho have been either extirpated or support only small remnant populations. Stormwater runoff in combination with stream channelization, passage limitations and stream cleanout have altered habitat in these systems (Clallam County 1993). Dams Hydroelectric development is limited in the region and has affected the Elwha River in the SJF'. Mainstem dams were constructed on river mile 4.9 in 1914, and river mile 12.0 in 1925. Both facilities were constructed without passage facilities. This action has prevented coho salmon from accessing 42.9 miles of mainstem and 32.05 miles of tributary habitat, much of which is pristine, located within Olympic National Park (FERC 1993). Arguably, construction of the dams on the Elwha River has impacted coho production in the SJF more than any other single human. activity. The Elwha River and its tributaries represent 21.90 of the total potential habitat area of the SJF based upon estimates of accessible stream habitat alone (Zilges 1977; James River II 1988) . Natural coho production in the Elwha is currently limited to a few side - channels and spring -fed floodplain streams in the lower river. A production coho hatchery has been constructed by the LEKT to supplement the run. Current Stock Status Although hampered by a lack of historical data, assessments of 4A penstock diversion is operated by the City of Port Angeles on Morse Creek. Water is diverted above an impassable wat`rfall for power generation and flows back to the system 0.5 below the fails. SBased upon 1987 low flow s=.-evs (FERC 1993). Thew estimates are considered cons ervacive of actual available tributary habitat. The survey methodoloav appears to have omitted sanificant reaches of small, low- gradient screams. 8 salmon stock status in the SJF have been made by McHenry et al. (1996) , Huntington et al. (1995), WDFW et al. (1994) , Nehlsen et al. (1993) and Lichatowich (1993a). These reports took somewhat differing approaches to the problem of defining individual stocks and stock status. For example, WDFW et al. (1994) assessed stock health in terms of how short -term changes in escapement related to currently available habitat quality. This approach led WDF et al. (1994) to conclude a healthy status for the following SJF coho stocks: Chimacum Creek, Salt Creek, and Hoko River. In contrast, McHenry et al. (1996) used definitions of stock health based upon life history strategies and minimum numbers required to prevent gene introgression (WDFW In Preparation) , and found no healthy coho stock in the SJF. Other authors reached similar conclusions (Huntington et al. 1995; Lichatowich 1993a; Nehlsen et al. 1993). TOTAL AVAILABLE STREAM HABITAT IN SJF Estimates of total stream habitat in the SJF have been made by Zillges (1977), who used the data in order to estimate coho smolt production. Recent surveys by the LEKT indicate that these estimates need to be updated, as there are significant errors with regard to habitat availability. An updated estimate based upon unpublished data provided by WDFW and the LEKT is given in Table 1. Because stream productivity varies greatly in the Su'F, we assigned a habitat quality index (poor, fair, good) to each watershed to reflect these differences. COHO SALMON LIFE HISTORY . Coho salmon are impacted by habitat alterations or losses at- various stages of their freshwater life history. Life history, patterns of coho salmon in SJF are not well documented, but are likely similar to coho salmon in other Washington streams. A generalized life history includes the migration of adults to them natal streams and spawning during October - January, egg deposition in gravel nests or "redds ", emergence of _-ry in spring (March -June! Table 1. Estimated accessible stream habitat to coho salmon b%.- 9 stream in the Strait of Juan de Zillges (1977); Unpublished Fuca Region, Washington. Source: Data WDFW & LEKT. Stream Accessib a Updated Watershed Length(mi) Length(mi) Production ( Zillges 1977) (LEKT \WDF) Rating Chimacum Creek 12.0 12.0 Tributaries 10.5 8.0 Poor Snow Creek 4.9 7.0 Fair Tributaries 1.3 3.0 Poor Salmon Creek 5.7 5.7 Fair Tributaries 6.5 6.5 Poor Discovery Bay Tribs. 6.9 6.9 Poor Jimmycomelately Ck. 1.9 Poor Sequim Bay Tribs. 6.0 1.9 Pair Dungeness River 18.8 8.7 18.3 Poor Tribs. to Dungeness 15.1 15.1 Poor Gray Wolf River 9.6 9.6 Poor Tribs. to Gray Wolf 2.8 2.3 Good McDonald Creek 5.2 S., Good Tributaries 2.5 4•. :air Siebert Creek 8.5 8.5 Fair Tributaries 2.0 3 Fair Bagley Creek 1'2 - 1.4 Fair Morse Creek Poor Pt. Angeles Streams .9 13.6 4.9 1.2 p oor Elwha River 4'9 Poor Colville Creek 1'S 4 4.9 Poor Salt Creek 6.5 9" 5 =air Tributaries 6.3 10.0- Good Whiskey Creek 4.3 Good Tributaries 4.0 1.0 2.J :air Field Creek Lvre River 2.8 2•� Pair Tributaries 4'0 2.3 Coca Murdock Creek 4 Good East Twin River 7 2.5 Good Tributaries 9.6 3 8.9 Good West Twin River 4.2 4.2 Good Tributaries 4'1 6 Good Deep Creek Good Tributaries, �:� Poor Joe Creek i.6 " c- Jim Creek 3.6 1 Fair - But' -er Creek 4. - Poor Pvsht River __ i6.3 - 14.E Fair Tributaries ailam River 24.3 2z " =air Gocc Tributaries 11.0 11. rater goko River 19.2 23 s 20 -; Good Tributaries 40.9 23.-* 43 - Sakiu River 3.3 3 Goca Tributaries i5 5 Occr J Olsen Creek Poor Jansen Creek 0 � Fa =r - F Rasmussen Creek 2': ai r Bull -man Creek } , - Tributari -s =a == Snow Creek Sail River Aaencv Creek _ -- Vilaae Creek Acc . `^.ainstem ACC. Trio. Length TOTAL �3;• ___.- and residence in stream h- abitats for z Year or more. Mizrati:;a 10 smolts enter saltwater the following spring and may remain in the vicinity of their river of origin for several months (Pearcy & Fisher 1986). The ocean migration period may last 18 months or more. Based upon tag recoveries, SJF coho migrate northward as far as Alaska. During this migration, coho are subjected to mixed - stock sport and commercial fisheries. Of significance to SJF coho, are troll and gill -net fisheries off the west coast Vancouver Island (WCVI) . There are currently no comprehensive studies of coho productivity in freshwater for any single watershed on the SJF. There are, however, aspects of their life history which have been studied in various watersheds. The approach in this section was to use the most current data available to describe factors limiting coho production in the SJF from a regional standpoint. Adult Migration Coho salmon accessing spawning areas face considerable migrational barriers in the SJF. Highway construction, logging roads, and hydroelectric dams are the most easily quantified impacts. Other potential hindrances include sediment filled channels, invasion of exotic aquatic and semi - aquatic plants, and loss of step -pool profile in upper channel networks from landslides. Such impacts are often subtle, and the severity effect may be dependent on interaction with flow regimes, run timing and run duration. As a result, many impacts are unquantified. H_ydroe l ectri c Dams From an areal perspective, construction of two mainst`m hydroelectric dams on the Elwha River has had the most significant impact on coho production in the SJF Yegion. The dams had a devastating impact on coho salmon of the Elwha, 'limiting production to just the lower 4.9 miles of the river. Subseauen—_ channel ization, fluctuations in flow regime, cessation of the 11 recruitment of gravel and woody debris has further degraded the productive capacity of the remaining habitat below the dams. The construction and operation of the Elwha dams and their effect on anadromous fish has a long history of controversy (Johnson 1995). During the 1980s, licensing of both facilities by the Federal Energy Regulatory Commission (FERC) had become extremely contentious and protracted, as various interest groups asserted legal, social and environmental arguments concerning the future of the dams. In an effort to resolve the dispute, Congress passed the Elwha River Ecosystem and Fisheries Restoration :tct (PL 102 -495. Signed in 1992 by President Bush, the act represents a negotiated settlement that protects the interests of the dam owners, municipal and industrial water users, Indian Tribes and Olympic National Park. The act authorizes the Secretary of the Interior to acquire and remove both projects if necessary to fully restore the ecosystem. The Secretary has concluded that removal of both dams is the only alternative that can achieve the goal of full restoration of the Elwha ecosystem and native anadromous fish (Doi et al. 1994; NPS 1995). The Elwha River is by far the largest (335 mil) river system in the SJF, and potentially the most productive. Over 800 of the drainage is located within the boundaries of Olympic National Park and is considered pristine habitat. It is estimated that 42.9 miles of mainstem and 32.0 miles of tributaries could be made accessible to coho salmon (DOI et al. 1995; James River II Potential coho production in a free - flowing Elwha River has been estimated at 248,964 smolts! year (FERC 1988) . In a f,a; lv producing SJF ( Elwha included) , this represents 35.9% of -"= total coho smolt production. In the currently degraded SJF, Elwha smolt Production would represent 6�.4% of the total smolt vie d. Urrbanizat4on 1Highwav Cons on Historic road construction practices, esnec; all•.• stY -an, crossings, involving the installat_on of culverts with de=_c -f-11S has .affected streams throuaheut the 3J . The state hian.,rav s%stenl 12 that circumnavigates the Olympic Peninsula (Highways 101 and 112) crosses every tributary to the SJF at least once. A recent inventory conducted by WDFW revealed the extent of the problem (Table 2). In addition to Highway crossings, urban development has affected coho salmon access to seven streams (Tumwater, Valley, Peabody, Ennis, White, Lees, and the East Fork Lees creek) in the Port Angeles urban area. These systems are affected by either impassible culverts, long reaches of underground culvert (up to 20001) or combinations of both. Access to these small tributaries is further compromised by the quality of available habitat which has been degraded by stormwater runoff, channelization, loss of woody debris, and estuary filling. Port Angeles harbor, once one of the largest natural estuaries in the SJF, has been completely altered to accommodate deep water shipping and industrial development. Sediment - Vegetation -Flow Interactions Accelerated sedimentation from a variety of land uses, can interact with channel conditions, vegetation, and flow to inhibit coho salmon access. In the eastern portion of the SJF flow characteristics are strongly influenced by the rainshadow of the Olympic Mountains. Rainfall in this region may be 10 -200 of that received in the western SJF. As a result, stream power is proportionally less. In low gradient channels, inorganic sediment may be stored in excess of the streams ability to transport. Increases in channel sediment are manifested by increases in channel bar area, channel meandering, and riffle crest height. During low flow periods (either natural and /or exacerbated by water :withdrawals), adult migration may be precluded. In the SJF, this has been documented for early spawning species such as summer chum (Salmon, Snow, Jimmycomelately Creeks) and fall chinook (Hoke River). Since coho salmon have a later run timing and presumab';f a higher probability of accessing spawning grounds during freshets, the magnitude of this problem is uncertain. However, this effect 13 Table 2. Inventory of impassable and partially passable culverts on Highway 101 and 112 on the Strait of Juan de Fuca, Washington. Source: WDFW (Unpublished Data). Stream Highway %Passable Comments Peabody 101 0 1.7 Miles Blocked 4.0 Miles Blocked 2.5 Miles Blocked White 101 0 Lees 101 0 Bagley 101 0 4.5 Miles Blocked Gierin 101 0 3 Miles Blocked by Tide Gate 3 Miles Blocked 3 Miles Blocked Johnson 101 0 Chicken Coop 101 0 Eagle 101 20 1 Mile Blocked Contractors 101 0 0.5 Mile Blocked 2.5 Miles Blocked 1.0 Mile Blocked 1.5 Mile Blocked Rasmussen 112 0 Butler 112 20 Joe 112 60 Nelson 112 0 0.5 Mile Blocked 2.5 Miles Blocked Limited Habitat Above Field 112 40 Whiskey 112 60 Salt Tributary 112 50 Bear Creek Salt Tributary 112 s0 Unnamed easterly_ Tributary Colville 112 <100 Tributaries to Clallam 112 0 (Tributaries Two Unnamed Tributary to Pysht 112 60 i Tributaries to Pvsht 112 0 ('Tributaries Two Unnamed may artificially narrow run timing. Colonization of stream channels 'ov the invasive e -c-4 . re_ canary grass h has bee :, documented as _z g (P alaris alterr._rlor ` , o�� significant problem for salmon on the O1�,-mn4c Peninsula (Personal 14 Communication, Jerry Gorsline, Washington Native Plant Society) . Reed canary grass can form dense mats, clogging channels and in some cases, impeding the passage of salmon. The inlet and outlet of Crocker Lake, which flows into Snow Creek, was recently cleared of reed canary grass by Jefferson County restoration groups. Infestations have also been reported in Ozette Lake, Chimacum Creek and tributaries to the Dungeness River. Human activities such as channelization, agriculture clearing and removal of riparian vegetation appear to create ideal conditions for reed canary grass blooms. Bahls & Rubin (1996) found that in Chimacum Creek, Reed Canary Grass respired carbon dioxide at night, depressing dissolved oxygen levels to in some instances <5 mg /L. In t ergravel Environment Coho salmon eggs may experience considerable natural mortality within gravel constructed redds, even in pristine habitats. Winter freshets can scour redds. Fine sediments can be mobilized and deposited into gravel interstitial spaces, smothering eggs or decreasing oxygen transportation. Redds may be constructed in marginal spawning areas during high flows, only to be de- watered when flows recede. Estimates of mortality due to natural incubation conditions are variable, depending upon rearing conditions, and range from Oo to 850 (Sandercock 1991). Human activities may further increase mortality rates for incubating eggs. For SJF coho, scour and increased sedimentation have been identified as the most critical factors limiting coho spawning success (Williams et al. 1975; McHenry et al. 1994; WDFW et a1. 1994; McHenry et al. 1996). Fine Sediment Elevated fine sediments ( <0.85 mm) in spaning gravel are known to decrease survival of developing salmonid eggs in the intergravel environment. In western Washington, fine sediment levels over 12-0s (by volume) are generally considered detrimental to 15 developing salmonid embryos (Peterson et al. 1993) . In the SJF, spawning gravel quality has been assessed in several watersheds, including the Sekiu, Pysht, Hoko and Clallam Rivers and Deep, Siebert, and McDonald Creeks. Results of theses studies generally show moderate to high levels of fine sediment (Table 4). A comparison of these values to estimates of survival obtained from egg survival studies in laboratory and natural environments indicates low predicted survival of eggs to the alevin and emergent fry stages in SJF streams. An in -situ test of coho salmon survival using egg- baskets and constructed redds (Burton et al. 1991) was conducted in the Hoko and Pysht River drainages between 1991 -93 (McHenry et al. 1994) . Results showed very low survival of coho eggs in the Pysht River (average of 2.80 to eyed stage and <10 to alevin) . Survival of coho eggs was higher, though variable (range 0 -580, average 25.7%) in the Hoko River. Survival was neither correlated with various measures of substrate quality (fine sediment, geometric mean particle size) , nor inter- gravel dissolved oxygen levels. However, at fine sediment levels >130, nearly 100% mortality was observed. This implies the existence of a threshold above which complete mortality occurs. The authors also observed significant bedload scour and fill events and hypothesized that channel stability was a significant factor affecting early life history survival. Channel Scour Processes Scour of spawning gravel has been frequently observed during peak flow events in Pacific Northwest streams (Lisle 1989; Nawa et al. 1990). Scour events are difficult to predict, and their effects on salmonids are related to a number of physical an biological factors including: 1) depth of egg deposition (related to body size) , 2) spawning timing, 3) magnitude and duration of flood flows, 4) size and quantity of bedload material, and channel characteristics (Schuett -Names e?- al. 109 =). lr, 16 Table 4. Fine sediment levels ( <0.85 mm) in spawning gravels reported for Strait of Juan de Fuca tributaries. Stream % Finea intensively managed watersheds, land management practices have resulted in conditions ideal for scour (i.e. increases in sediment yield, loss of LWD, increases in peak flow runoff). In the SJF, data on channel bed scour has been collected on the Dungeness and Pysht Rivers. During the winter of 1994 -95, Orsborn & Ralph (1995) placed 29 scour monitors (chains) at 16 locations in the mainstem Dungeness River. Twelve of the scour chains were placed adjacent to recently constructed pink and Chinook salmon redds. As part of a chinook restoration project, each chinook redd was later pumped of alevins, in order to collect fry for a captive brood -stock program. Of the twelve redds where monitors were installed, five showed evidence of substantial scour (4 -40 cm), while four were buried under bedload. Subsequent information provided by the Washington Department of Fish and Wildlife (WDFW) confirmed that only one chinook redd yielded anv_ viable alevins. Orsborn & Ralph (1995) concluded: 17 "chinook and pink salmon redds in the lower 10.8 miles are largely unsuccessful because the locations chosen for redd construction appear to scour deeply at even moderate flow events" Because coho salmon utilize both mainstems and tributaries for spawning, a discussion of channel bed stability in small streams is pertinent. Between 1989 -91, cross - sectional profile, thalweg depth and scour monitors were installed at 9 sites in the Pysht River. Sites selected included both mainstem (6) and tributaries (3). Results showed significant bedload movements: aggradation or degradation of >0.33 m was observed at 15 of 27 sites (Unpublished Data, University of Washington). Schuett -Names et al. (1995), based upon literature values, recommended that for medium - bodied salmon (such as coho), the mean egg pocket depth averaged 0.16 m. Similar bed instability was observed throughout the Hoko River basin during the early life history studies conducted by McHenry et al. (1994) . Winter Temperature Altered thermal regimes have been documented in numerous SJF streams during summer and are suspected in winter. Long -term studies in low elevation coastal watersheds of British Columbia have shown that elevated winter temperatures are an outcome of intensive logging (Hartman & Scrivener 1990) . Elevated winter temperatures has been shown to advance embryo development, resulting in earlier emergence timing, and ultimately earlier smolt outmigration. In the Carnation Creek, British Columbia study, coho reached a counting fence about two weeks earlier after the watershed was logged. Holtby (1988) found that the marine survival of early migrating smolts was reduced following loaaina presumably , p because of asynchronous timing between marine food resources and coho smolts. Although advancement of out«igration timing has not been reported in the SJF (or elsewhere in the Pacific Northwest), this effect may be significant, as biotic and abiotic conditions _D:: SJF watersheds are very similar to those of Carnation Cree4_. W First: Year Parr Coho fry that successfully emerge from the intergravel environment in spring, seek refuge in stream margin and backwater habitats. As coho parr grow larger, they typically search for pool habitats which contain suitable depth, cover and flow conditions. Flow and pool area are important factors regulating survival for summer parr. Quantity of flow directly limits available stream area for coho rearing. This relationship is particularly strong for Puget Sound coho stocks, as long -term smolt trapping has shown that smolt production is positively correlated with indices of summer flow. Coho parr are strongly territorial and aggressively defend favorable rearing sites. Pool size appears to be more critical than pool type during summer. In coastal Oregon streams, coho appear to utilize almost all pool types equally during the summer (Nickelson et al. 1992). In forested streams of the Pacific Northwest, pool forming processes are dominated by the interaction of hydrologic forces and channel obstructions. A constant supply of Large Woody Debris (LWD) is particularly important for the maintenance of pool features used by coho salmon. In unmanaged streams, volumes of down woody debris may be prodigious. Human activities, particularly logging, cedar salvage and channelization have dramatically altered the dynamic balance of LWD in streams (NRC 1995). The original old growth coniferous forests have been eliminated from SJF watersheds. Riparian buffers, have only recently been implemented (1982) on state and private timberlands in Washington State. As a result, every watersheds in the SJF has been logged without any buffers at least once. Current state riparian standards do not provide adequate recruitment of LWD. Not only have the overall volumes of instream wood been reduced, but the composition has been converted from decay resistant coniferous species to more fragile deciduous species. Historic logging practices, in combination with a lack, of 19 reforestation, have left a legacy of riparian forests dominated by early successional species such as red alder (Alnus rubra). Pre - settlement riparian forests on the SJF were characterized by relatively sparse densities (20 -50 trees /acre) of large conifers (Pentec 1995). Analysis of current riparian conditions on the S.F. Pysht (Unpublished Data, LEKT) and Hoko Rivers (Pentec 1995) showed that 98% and 930, respectively, of the riparian forest was dominated by red alder. These stands are characterized by high densities (100 -400 trees /acre) of small - diameter trees, often in association with dense under - stories of brush (i.e. salmonberry), that may preclude or delay the development of late - successional vegetation communities. Probably one of the greatest changes wrought on Pacific Northwest streams by man's activities (and the most significant for coho salmon) is the systematic loss of LWD (NRC 1995). Recent research has elucidated the importance of LWD for habitat formation, sediment storage and floodplain processes (Bisson et al. 1987) . In streams draining urlogged old- growth forests, the volume of down debris interacting with stream channels was large (Maser et al. 1988). Grette (1985) found that in unlogged streams of the Olympic Peninsula LWD averaged 89 m3 /100m of stream. Peterson et al. (1993) recommended debris loading of 2 -4 pieces /channel width for western Washington streams. In SJF streams, LWD levels are typically very low (Table 5) , and continue to decline. In a survey of 28 Olympic Peninsula streams (including 6 in the SJF) between 1982 -93, McHenry et a_'. (In Preparation) found that total LWD volumes had declined by an average of 25.90 (Figure ?1. The average volume of LWD in 1993 was 49.80 less than the volume found in seven unmanaged streams during 1983. LWD depl`t;cn has implications to stream stabili:_ (scour issues) , summer and :.Tinter rearina. 20 120 100 80 a 60 40 20 140 120 100 a 80 >' 60 40 20 1_ n Total Volume of LNVD 1982 Group U = Unlogged Y = Young Sec: Growth ■ O M = Middle -Aged Sec. -Growth O = Old Sec: Growth 1993 > 1982 / 1982 = 1993 ■/E 0 ■ �1 O ■O %U ■O ■Y N 0 N O n a M t 'i ■Y 0 ■h0U a hi 1982 > 1993 ■U ■U 40 60 80 100 120 Year 1982 no Total Number of LIVD Pieces a n 1993 > I982 ■ ■Y ■U M O y no U 'Y ■ R1 no ■nl ■Y ■h'Y ■U I`i 1982 > 1993 IU no ■U no 1982 Group U = Unlogged Y = Young Sec. -Growth M = Middle-Aged Sec. -Growth O =Old Sec.-Growth :.v 40 60 80 100 120 Year 1982 Figure 2. Decadal changes in LWD volume for 28 Olympic Peninsula streams, 1982 -1993. From McHenry et al. (in Preparation). 21 Table 5. Measures of Large Woody Debris loading found in watersheds of the Strait of Juan de Fuca, Washington. Stream # Pieces/ Volume/ inn m Inn _ From the standpoint of salmon ecology, current riparian forests function differently from the conditions encountered in old- growth forests. Contribution of nutrients, thermal rea4mes, and interactions between floodplain and river habitats have been altered over historic conditions. For example, research has shoven that nutrient inputs to streams bordering old growth forests are different from forests that have been logged (Bilbv & Bisson The implications of these changes to species such as coho sa-Imcn that have changing spatial and temporal habitat requirements are not fully understood at this time. Summer rearing habitat for first year parr has been measured in several SJF watersheds (Table 6 ) . Indices of summer a habitat quality indicate that summer rearing areas are gen =-ral - low, though variable. In y N e systems that ha-� e been hea���ily 22 S.F. Pysht River 27.4 30.6 LEKT (Unpub. Data) Deep Creek 7.5 11.4 LEKT (Unpub. Data) Hoko River 13.5 - -- Pentec (1995) Little Hoko River 9.1 10.3 LEKT (Unpub. Data) Chimacum Creek 4.3 - -- Bahls & Rubin (1996) PNPTC (Unpub. Data) Snow Creek 6.2 - -- Dungeness River 2.8 - -- Orsborn & Ralph (1994) LEKT (Unpub. Data) Siebert Creek 3.1 - -- Pysht River 18.0 9.8 LEKT (Unpub. Data) Pysht Tributaries 22.8 11.8 LEKT (Unpub. Data) AVERAGE 11.5 14.8 From the standpoint of salmon ecology, current riparian forests function differently from the conditions encountered in old- growth forests. Contribution of nutrients, thermal rea4mes, and interactions between floodplain and river habitats have been altered over historic conditions. For example, research has shoven that nutrient inputs to streams bordering old growth forests are different from forests that have been logged (Bilbv & Bisson The implications of these changes to species such as coho sa-Imcn that have changing spatial and temporal habitat requirements are not fully understood at this time. Summer rearing habitat for first year parr has been measured in several SJF watersheds (Table 6 ) . Indices of summer a habitat quality indicate that summer rearing areas are gen =-ral - low, though variable. In y N e systems that ha-� e been hea���ily 22 Table 6. Indices of first year parr summer rearing habitat for coho salmon, Strait of Juan de Fuca Region, Washington. STREAM MEAN MEAN DATA POOL POOL SOURCES AREA DEPTH M (m) Hoko River 39.2 0.53 Pentech (1995) Hoko Tributaries 38.7 0.63 Elwha Klallam Tribe SF Pysht River Pysht Tributaries 42.8 0.72 42.7 0.63 (Unpublished Data) McHenry et al. (1995) Deep Creek 46.6 0.70 Morse Creek 12.0 - - -- Peninsula College (Unpublished Data) Snow Creek 37.5 0.44 Rowse (In Preparation) Siebert Creek 29.6 - - -- LEKT (Upub. Data) McDonald Creek 34.5 - - -- PNPTC (1996) JimmyComeLately Ck. 10.4 - - -- Donald (1993) Chimacum Creek 49.6 0.51 Bahls & Rubin (1996) AVERAGE 34.8 0.59 channelized or managed for agricultural uses, pool habitat has largely been eliminated (i.e. Morse & Chimacum creeks) . In streams draining watersheds managed for forestry, pool habitat persists, though below levels found in unmanaged forests. Peterson et al. (1993) recommended a threshold of 50o for pool habitat in low - gradient western Washington streams. Pool quality in SJF streams indicates relatively homogenous conditions: available pools are typically shallow ( <1.0 m) depth and lacking in complex cover. These conditions aPPear to be related to pool filling as a result of accelerated sedimentation. Shaw (1903) found through a sediment budget analysis in Green Creek, a tributary of the Pysht River, that 85 -940 of the sediments resulting from mass- wasting in the 1960 -1970s were still stored within the valley floor. Green Creek has.downcut through these deposits (0.5 -4.0 m depth) and is 23 currently releasing fine sediments to the channel at a rate of 131 M'/km2/yr. This represents a persistent source of fine sediment. Since summer carrying capacity is related to pool volume, loss of depth represents a significant loss of available space. The quality of summer rearing habitat is also affected by thermal regime. Research indicates that summer temperatures between 12 -14 °C are optimal for juvenile coho (Brett 1956; Reiser 1979). Temperatures exceeding 20 -25 °C may cause both direct and indirect mortality for coho salmon (Brett 1956). Stream temperatures are affected by both natural conditions (average watershed elevation, aspect, groundwater influences) and anthropomorphic alterations (removal of riparian vegetation, channelization, stream widening). Temperature regime in SJF streams is highly stream, and even reach specific, making generalizations difficult. In streams draining lands managed for agriculture, thermal regime may be substantially altered. Agricultural practices such as riparian clearing, stream channelization, and surface and groundwater extraction, may exacerbate flow conditions during summer. Data collected from lowland, agricultural streams in the SJF show that summer temperatures may reach levels detrimental to juvenile coho salmon. During 5 years of summer stream temperature monitoring in Chimacum Creek, Bahls & Rubin (1996) consistently found summer temperatures ranging from 14- 21.5 °C. Jones & Stokes (1991) reported the maximum temperatures for Snow Creek between June and September as 19.3, 20.3, 20.3, and 18.1 °C. Data from lowland western SJF streams indicates similar summer temperature patterns (Unpublished Data, Lower Elwha Klallam Tribe). Win cer Parr With the onset of winter, juvenile coho migrate dorr.stream seeking habitats that provide refuge from winter freshets.n contrast to summer habitat, winter habitat types vary in their ability to support coho parr (Nickelson et al. 1992), Off - Channel 24 habitats such as side - channels, alcoves, wall -based channels, and beaver dams represent ideal areas for winter rearing from a metabolic standpoint: Off- channel habitats have high food production, and require less energy expenditure than other habitats. Because off - channel areas are usually associated with low gradient alluvial valleys, they have also been subjected to a proportionally higher level of human development. These areas were typically the first to be cleared for agriculture, human settlement and subsequent flood control efforts (Seddell & Froggart 1984). In the SJF, loss of over - wintering habitat is a significant problem for coho salmon. Historic losses are only available for Chimacum Creek, where Bahls & Rubin (1996) estimated that 970 of the over - wintering rearing capability for coho had been lost since pre - settlement times (1850s). Estimates of available present day over - wintering habitat in other SJF streams indicates extremely low numbers (Table 7). Disconnection of flood plain habitats from mainstem rivers is probably one of the most serious cumulative effects of man's activities on fluvial systems. Because of the importance of these habitats to coho salmon life history, protection of remaining intact floodplains and restoration of disconnected floodplain environments should be a high priority. Beechie et al. (1995) found that loss of off - channel habitat, primarily because of agricultural development was the most significant factor limiting coho production in the Skagit River. Smo1 tif.ication- Migration to Marine Wafers Coho salmon that survive winter rearing conditions, undergo the physiological preparation for the transition from freshwater to marine environments (smoltification) . Smoltification is triggered by changes in photoperiod and temperature. During this period coho salmon are sensitive to chemicals commonly applied across the landscape such as insecticides and herbicides. In the SJF, 25 Table 7. Estimates of currently available off - channel habitat as a percentage of total available habitat types in Strait of Juan de Fuca Tributarieq_ wa�r,;nnr�r `Aerial aAerial photo` estimates STREAM AVAILABLE COMMENTS estimates estimates OFF - CHANNEL HABITAT ( %) Snow Creek 0.7a Channelized and cleared for agriculture. Dungeness River Lower 10.0 miles have been - -- cleared and heavily diked (4.5 miles) . Morse Creek 0.0° Lower 2.0 miles have been cleared and Channelized. Elwha River 10.0` Dam construction, diking and channelization has isolated mainstem from floodplain. Deep Creek 2.3' Massive sedimentation from logging, subsequent channel widening have isolated floodplain from mainstem. Pysht River <10.0e Highway construction along lower 8.5 miles of mainstem. Little Hoko River 3.6= Channel downcutting to 4 m in response to flood waves from dam -break landslides. Sekiu River <5.09 Logging roads constructed through lower 4.0 miles of mainstem. TT 3LP_S 1'P1 1 A oY' �1 � A[1 '1 nt •____ `Aerial aAerial photo` estimates dMcHenryJet�alac19g5enlnsuia College 3Aeriai photo photo estimates estimates (Unpublished Data, Elwha Kiallam Tribe herbicides such as glyphophospate are commonly used during spring to suppress deciduous brush in recently planted clearcuts'. Although herbicides are generally carefully applied, according to manufacturers recommendations and current forest practice rules, 'This may be a problem throughout the range of coho salmon. Along the entire Pacific Coast coho habitat an�est_ mated 150,O0p pounds of 2,4 -0 were applied in 198% ;Pair 1992). 26 little is known about the potential acute or chronic affects of such practices on juvenile coho (Grier et al. 1994). Coho Salmon Life History Models To facilitate the discussion of impacts to coho salmon during freshwater residence, two simplified models that describe mortality factors for coho salmon in freshwater were developed. Sufficient data has been collected from a SJF streams to develop estimates of the effects of habitat degradation on coho salmon. Two systems, the Pysht River and Chimacum Creek, represent examples of the different types of land management practices that occur in the SJF. The case history for Chimacum is based upon a habitat restoration assessment conducted by Bahls & Rubin (1996). The model for the Pysht River is based upon basin wide habitat surveys conducted in 1993 (Unpublished Data, Elwha Klallam Tribe), assessments of intergravel survival (McHenry et al. 1994) and streambed scour (Unpublished Data, University of Washington- Center for Streamside Studies) . Paradoxically, although these streams represent extremes of hydrologic conditions encountered in the SJF, they share similar impacts with regards to loss of over - wintering habitat. Conditions encountered in Chimacum Creek are shared by roughly 2695 of the stream mileage in the SJF, while those of the Pysht represent 620 of the total. The remainder of streams in the SJF (14') are impacted by urbanization. Chimacum Creek Draining an area of 37 mil, Chimacum Creek is the largest and most productive drainage in the Admiralty Inlet area of the eastern SJF. The system is "Y" shaped, with the East and West Forks flowing together to form the mainstem at river mile 3.0. Chimac'.:m Creek is located in the Olympic rainshadew and recei• ✓es about 22" /year of precipitation. All land adjacent to Chimacum Creek is Privately owned. Land uses include forestry in headwaters areas and extensive agricultural development along low - gradient 27 alluviated portions of the mainstem and lower East and West Forks. Chimacum Creek historically supported runs of coho, summer chum salmon, and steelhead. Summer chum are extirpated from the basin (Lichatowich 1993a). Coho salmon are considered threatened with a total run size of <250 fish (Bahls & Rubin 1996; McHenry et al. 1996). Historical estimates of coho escapement, through anecdotal observations are estimated at 1000 -2000 fish (Bahls & Rubin 1996) . Although the Chimacum valley historically supported cedar and spruce swamps, wet prairies and beaver ponds, most of the lowlands were drained and converted to pasture at or before the turn of the century. The area was settled beginning in the 1850's. Initial development consisted primarily of logging and clearing in the valley bottoms for homesteads. The surrounding hillsides were then logged and roaded. Beginning in 1919, main stem and tributarvv streams in the valley were channelized to drain agricultural land. Ditching continued into the 1970's and reed canary grass was widely introduced as forage for livestock. These activities dramatically changed the habitat conditions in Chimacum Creek (Figure 5) . Basin -wide surveys conducted by biologists from the Port Gamble S'Klallam Tribe quantified the extent of historical changes in the basin and related these to coho freshwater life history: * A 12% loss of spawning habitat due to impassable culverts. General degradation of remaining spawning habitat due to sedimentation. * Summer rearing areas have been reduced by 945.. Water temperatures between 16 -21 °C were recorded in all non - forested reaches; dissolved oxygen levels <5 mg /L were recorded at four sites. Because of degraded conditions, summer parr were generally absent from extensive reaches of the West and East Forks. Clearing, ditching, groundwater extraction, and elimination of beaver have likely altered summer flow patterns to the detriment of salmonids * Because of these development activities, winter rearing areas have been reduced by 97a. The majoritr of the drainage has few or no instream LWD. Flooding may, strand juvenile coho i:: pastures after flood water recede. 28 H r �\ C f��rlA «/III C/?E"EK clir,-,a /6V Figure 3. Changes in stream habitat in Chimacum Creek.bet:aeen circa 1800 and 1995. From Bahis & Rubin (1096). 29 Pysht River The Pysht River is a moderate sized drainage (30,000 acres) located in the western SJF. The Pysht mainstem is 16.5 miles long and receives drainage from several significant tributaries. Precipitation patterns in the Pysht are more similar to those found in the coastal Olympics. Although historic estimates of coho salmon production are unavailable, anecdotal information supports the contention that the Pysht supported a sizable and diverse salmon population. In contrast to Chimacum Creek the Pysht is managed almost entirely for commercial timber production. The basin was historically occupied by bands of S'Klallam peoples who lived permanently at a settlement near the mouth of the river. The area was first settled by Euro- Americans in the 1.870s. Commercial timber operators followed, and by 1915 full -scale railroad logging operations were initiated. By 1925, the majority of old growth accessible to railroad logging had been harvested. After World War II, truck logging, which allowed access to upper portions of the drainage network proliferated. Intensive clear- cutting peaked in the late forties and again in the late 1980s7. Logging has significantly affected salmon habitat in the Pysht River. Sediment yield as a result of intensive roading, and landsliding has increased (Benda 1993; Shaw 1993) . Channel conditions, particularly as affected by loss of LWD and increased sediment yield have been degraded. Thermal regime has been altered over the conditions coho salmon evolved to in the basin. Despite these problems, the situation i s s not hopeless. Recent improvements in forest practices, ongoing restoration projects, combined with the heightened concern of one large landowner (and departure of another) provide hope over the long term. The Pysht still contains 'Clear- cutting in the 198Cs was driven by corporate financina debts ( "junk bond forestry "). The forestry practices used in the Pysht inspired controversy that helped persuade the Washington to revise its Forest Pract -ce Code. state of 30 habitat characteristics critical to coho salmon, namely available over- wintering habitat areas and an intact estuary. Pysht River Coho Life History Simulation Adults -We assumed a hypothetical escapement of 1000 adults to the Pysht River. Actual estimates of total escapement to the Pysht are unavailable, but are likely less than 1000 fish annually. Based upon a 1:1 sex ratios and an average fecundity of 2,500 eggs /female for Washington origin coho (Salo & Bayliff 1958; Fraser et al. 1983), a yield of 1,250,000 eggs is expected in spawning habitats of the Pysht River. Egg Survival -A complete assessment of spawning gravel quality has been conducted in the Pysht River. Based upon 135 McNeil Core samples collected in mainstem and tributary spawning areas, McHenry et al. (1993) found that the levels of fine sediment <0.85 mm averaged 17.30. Increasing levels of fine sediment in intergravel environments are known to be detrimental to developing salmonid eggs. In western Washington watersheds, Peterson et al. (1993) has shown that fine sediment levels <126 are considered favorable, levels between 12 -176 are considered moderately degraded, and levels exceeding 176 are harmful. In order to assess expected egg survival, we compared average fine sediment values in the Pysht to coho egg survival curves developed from both laboratory and in -situ studies in the Northwest. From these curves a range of survival values was determined (9 -550). Applying these expected survival rates to the total egg deposition yields three estimates (low, medium, high) of total fry emergence (112,500, 312,500, 687,500). Survival to emergence is also affected by channel scour processes in the SJF streams, and should be considered an additive mortality factor. Stream bed scour is affected by the relationship 'The literature indicates that sex ratios for coho are variable and are often skewed by large numbers of jacks. Literature values for ratio of male to female coho from Alaska and British Columbia ranged from 0.9:1 to 3.1 :l (Sandercock 1991) 31 between peak flow, sediment size and yield, and channel characteristics (gradient, roughness). Although these processes have been widely observed, its relationship to land management practices has not been quantified. However, there appears to be a relationship between scour and streams that are channelized (Orsborn & Ralph 1995), depleted of woody debris (Smith et al. 1993) or heavily aggraded with sediment (Schuett -Names et al. 1995) . These are conditions common to many streams in the SJF. Morrill et al. (In Preparation) concluded that bankfull discharges of a 1 -2 year recurrence rate caused significant mortality to developing coho embryos in the Pysht River. Based on these observations, we have added a mortality factor of 0.5. Applying this factor to the range of survival values expected from fine sediment mortality yields 56,250, 156,250, and 343,750 total fry. Spring -Se=er Parr -Newly emerged coho fry (length around 30 mm) typically emigrate to quiescent water areas such as stream margins, and backwater areas and begin feeding. By early summer, fry disperse to stream habitats, particularly pools, where they compete for food and space. Competition for available space is intense, with the most favorable habitats typically being occupied by the largest individuals. Displaced fish are forced to locate new territories or are eventually displaced downstream, and in some cases to the estuary (Mason & Chapman 1965). Predation is also a mortality factor. The abundance of coho in a stream during summer is directly related to the number and complexity of territories available (Larkin 1977; Nickleson et al. 1992). At current seeding levels in the Pysht River, summer rearing habitat for coho does not appear to be limiting. A basin -wide survey of stream habitat indicates that approximately 70,000 and 123,000 m= of tributar�,,and mainstem summer rearing habitat, respectively, are available for summer parr in the Pysht (Unpublished Data, Lower E1wha Klallam Tribe). Despite the basin -wide depletion of LWD, a variety of pool types (largely formed by bed features) are available for rearing. A combined spring- summer survi %ral factor of .65 :aas assumed for the 32 Pysht River (Reeves et al. 1989) . Applying this factor to potential fry survival yielded a range of 36,562, 101,562, and 223,437 total summer parr. Assuming even distribution over available habitats, juvenile coho densities of o.18/m, (low), 0.52/m2 (medium) , and 1.15 /m= (high) could be expected. The low range compares favorably with actual coho summer parr densities found in Pysht River tributaries (Unpublished Data, Elwha Klallam Tribe). It should be noted that summer habitat in the Pysht River becomes fully utilized at 965,000 fry (assuming a density of 5.0 fish /m= of pool area) . Based upon assumptions of intergravel mortality factors, as few as 2,500 adults (1:1 sex ratio) could fully utilize ( "seed ") the available summer rearing habitat in the Pysht River. Winter Parr -With the onset of winter and decreasing water temperatures, juvenile coho salmon redistribute into deeper pools (particularly with complex cover) and off - channel habitats (Bustard & Narver 1975; Scarlett & Cederholm 1984). Movement is generally in a downstream direction (in response to freshets) and considerable distances may be involved. Coho streams with the best over - wintering habitat typically contain spring -fed ponds, beaver ponds, or side - channels (Narver 1978; Peterson 1980). Coho that are able to access such wintering habitats typically survive at very high rates. For example, survival of winter parr in Carnation Creek, British Columbia (prior to logging) exceeded 670 (Tscpalinski & Hartman 1983) In western Washington streams, over - wintering habitat has been substantially altered. Estimates of over - winter survival for winter parr in western Washington and Oregon generally average less than 350 (Bustard & Narver 1975; Reeves et al. 1989). In the SiF, over- winter survival may be much 33 higher. In the Pysht River potential over - winter habitat' was estimated at 15,340 m= for tributaries and 38,275 mz for lower mainstem areas. There is however, no data available to assess actual survival in these areas. Based upon the literature values for over - wintering habitat, we assigned an average survival rate of 0.325. Applying this to the total expected summer parr yields a range of 11,882, 33,007, 78,202 potential coho smolts. Smolt to Adult Survival- Estimates of smolt to adult (marine) survival are available from Snow Creek in the eastern SJF (Lestelle et al. 1993), and from Big Beef Creek, a tributary to Hood Canal (Quinn 1994). These data show some disturbing trends in marine survival of coho salmon. In Big Beef Creek for the years 1976- 1978, (Quinn 1994) found that survival to adult averaged 20.106. More recent estimates for the period 1973 -1989 show that marine survival averaged 22.9. (Lestelle et al. 1993). There are indications that marine survival rates for SJF coho are much lower than those for Puget Sound stocks. For Snow Creek, only combined ocean mortality (natural and harvest) estimates are available (Lestelle et al. 1993) . For the period 1976 to 1989, smolt to adult return rates averaged 3.9. (range 1.1 to 7.7). The authors observed a declining rate of return, that was not attributable to any one specific cause. We applied a marine survival of 15.5. to the range of smolt yields and obtained estimates of 1,842, 5,116, 12,121, adult coho. Survival to Escapement -The final mortality factor are those associated with commercial fisheries. Coho salmon from the SJF are subjected to fisheries in Alaska, British Columbia, coastal Washington, and Puget Sound. Exploitation rates have been hiah: Lestelle et al. (1993) analyzed exc?oitation rates for Big Beer 9Over- wintering habitat is def_ned as pools formed by large wood (logs, roots, or jams) with a residual dep::h >i m, side - channels or off - channel ponds. 34 Creek coho for the years 1975 -1988 and found a average harvest of 77.40 (range 55.6- 90.7 %). The authors reported no obvious trends in exploitation rate over time. For stocks of the western SJF, there is considerable concern about increasing exploitation rates in specific fisheries. Coded -wire tag recoveries for the period between 1986 -1990 showed an increasing exploitation rate in west coast Vancouver Island troll fisheries. Assuming an average exploitation rate of 65% yields a potential escapement range of 644, 1,790, and 4,242 fish. 35 CONCLUSIONS The Strait of Juan de Fuca covers an area of approximately 1500 mil and includes some 48 independent watersheds that support coho, ranging in size from <10 mil to >300 mil. Hydrologic conditions due to differences in watershed size, elevation and position with regards to the rainshadow of the Olympic Mountains, are highly variable in the region. Coho salmon populations in the SJF are significantly impacted by degradation of freshwater habitats. Low marine survival rates and high interception rates also hamper stock productivity. Based upon the minimum numbers of fish necessary to prevent genetic introgression (WDFW In Preparation) we were unable to identify any healthy stocks of coho in the region. This finding is directly in contrast to other stock assessment studies (WDFW et al. 1994). Failure to meet Escapement Goals SJF coho stocks have consistently failed to meet escapement objectives established by fisheries managers. One possible outcome of the PFMC stock review process is an assessment of escapement objectives in the SJF. A combination of low freshwater survival rates, marine survival rates, and high exploitation rates (in mixed -stock fisheries) has allowed SJF coho to only attain a maximum of 7001 of the escapement goal only once since 1983. The cumulative effects of these factors create some of the most difficult management conditions for any wild coho stock in Washington. Under these conditions, discussions of redefining escapement goals may be irrelevant, unless all freshwater and marine survival factors can be improved simultaneously. Substantial evidence exists that increased mortality occurs at all stages of coho life history in the SJF. Impacts to freshwater habitat are agruably amongst the greatest of any region in Washington. The two largest and potentially most productive watersheds in the region have been heavily hydro- modified.and are currently managed primarily for hatchery production. Approximately- 620 of regions watersheds have been subjected to intensive logging 36 operations and are highly destablized. Streams in the eastern SJF are influenced by rainshadow cast by the Olympic Mountains and have flow conditions that are similar to those of streams at the extreme southern edge of their range (N. California /Southern Oregon) . Habitat degradation caused by logging,- urbanization and agricultural activities in combination with these flow conditions is severe enough to cause local extirpation of coho stocks. Recent data collected by the WDFW and LEKT indicates that this may have already occurred in a number of eastern SJF streams. Elwha River Construction of hydroelectric dams without passage facilities on the Elwha River at river mile 4.9 and 12.0, effectively eliminated 35.9 miles of mainstem and some 42.9 miles of tributary from production. Much of this habitat is considered pristine, located within the boundaries of Olympic National Park. Conditions in the lower 4.9 miles of the Elwha (below Elwha Dam) are severely degraded due to cumulative effects of channelization and loss of gravel recruitment. The Elwha River is currently managed almost entirely for hatchery coho production. Potential smolt coho production from a restored Elwha River represents between 36 -650 of the total from the SJF region as a whole. Restoration of the Elwha River represents the single best opportunity for SJF coho stocks as a whole. Dungeness River Habitat conditions in the Dungeness River are so severely degraded that only limited natural production. occurs. The lower 11 miles of the Dungeness River has been extensively channelized and systematically cleared of large woody debris. Water withdrawals Place additional strain on salmon populations located in marginal flow conditions. Surface water withdrawals on the Dungeness River currently average about 6001 of the river's natural stream flow. A total of 579 cfs could be legally removed from the Dungeness River, even though the summer low flows average around 200 cfs. Over 400 37 miles of irrigation canal have been constructed in the Dungeness Valley to deliver river water for agricultural purposes. Many of these diversion are either unscreened or ineffectively screened. A healthy Dungeness River represents 18.50 of the total coho smolt yield of the SJF. The Dungeness should also be a high priority for restoration in the region. Western Strait of Juan de Fuca The western SJF is located within one of the highest runoff yield zones in western Washington. Hydrologic stress on biological systems are extremely high in the western SJF, including: 1) Soil types with high water delivery potential, 2) High drainage densities, 3) High road densities, 4) Destabilized channel networks from LWD depletion, accelerated sediment yield. In the western SJF, impacts to freshwater habitat are primarily limited to those associated with 80 years of intensive timber harvest, and in some cases hydromodifications on lower mainstem rivers. The region's old- growth forests have been rapidly converted to tree farms, as Olympic National Park affords little protection to SJF drainages. Historic management practices have left watershed conditions typically destablized with high road densities, accelerated rates of mass wasting, and altered riparian communities. Past management practices such as stream cleanouts, cedar salvage and channelization have further deteriorated habitat conditions. Because of these conditions, current forest practices rules, though improved, may be inadequate to allow watershed recovery in the western SJF. We modelled mortality factors for Pysht river coho throughout their life history (Table 8) and found that mortality factors in the winter were significant. Losses in the intergravel environment due to elevated fine sediment levels and scour may exceed 751- during some years. Over- wintering mortalit, is also suspected to 38 Table 8. Summary of life history impacts to coho salmon of the Pysht River, Washington, a representative stream of the western Strait of Juan de Fuca. Based upon escapement of 1000 fish. Life Stacre 17; Ok D—A.., A be high (7011), as habitat has been lost to diking, road construction and LWD depletion. At current seeding levels summer habitat is adequate, however at marginally higher escapements, this habitat type may also become limiting. Eastern Strait of Juan de Fuca In the eastern SJF, the cumulative effects of agricultural development, water withdrawals, channelization and urbanization in combination with high exploitation rates and limited natural flow regimes have driven coho populations perilously low. Recent survey data indicates that coho populations in the eastern SJF are in serious jeopardy: many stocks are declining at an alarming rate. Water production per unit area in this portion of the SJF is more similar to that of southern Oregon or Northern California. These conditions present a higher risk of local extinction in the SJF, similar to those seen for salmon stocks at the geographical extremes of their natural distribution. 39 Spawning 56,250- 343,750 Survival a function of gravel quality /channel stability Summer 36,562- 223,437 Density estimates are a Rearing function of flows and pool quality, which is a function of LWD cover and residual pool depths Winter 11,882- 78,202 Loss of off- channel habitat Rearing and subsequent alteration in peak flows Smolt to Adult 1,842- 12,121 Marine survival much lower for SJF stocks than those of Puget Sound Survival to 644- 2,242 Increasing rates of Escapement exploitation particularly off Vancouver Island be high (7011), as habitat has been lost to diking, road construction and LWD depletion. At current seeding levels summer habitat is adequate, however at marginally higher escapements, this habitat type may also become limiting. Eastern Strait of Juan de Fuca In the eastern SJF, the cumulative effects of agricultural development, water withdrawals, channelization and urbanization in combination with high exploitation rates and limited natural flow regimes have driven coho populations perilously low. Recent survey data indicates that coho populations in the eastern SJF are in serious jeopardy: many stocks are declining at an alarming rate. Water production per unit area in this portion of the SJF is more similar to that of southern Oregon or Northern California. These conditions present a higher risk of local extinction in the SJF, similar to those seen for salmon stocks at the geographical extremes of their natural distribution. 39 A historic assessment of Chimacum Creek showed that destruction of riparian forests, associated wetlands, extirpation of beaver, and channelization has resulted in a 94 and 970 loss of summer and winter habitat for coho salmon, respectively. These losses parallel declines in productivity for coho in the basin. The conditions described in Chimacum are representative of approximately 260 of the stream mileage in the SJF. 40 LITERATURE CITED Bahls, P., and J. Rubin. 1996. Chimacum watershed coho salmon habitat restoration assessment. Unpublished Report, Port Gamble S'Klallam Tribe, Kingston, Washington. Beechie, T., E. Beamer, and L. Wasserman. 1994. Estimating coho salmon rearing habitat and smolt production losses in a large river basin, and implications for habitat restoration. North American Journal of Fisheries Management 14:797 -811. Benda, L. 1993. Geomorphic analysis of the South Fork of Green Creek. Unpublished Report, Washington Department of Natural Resources, Olympic Region, Forks. Benda, L., T.J. Beechie, A. Johnson, and R.C. Wissmar. 1992. Morphology and evolution of salmonid habitats in a recently deglaciated river basin, Washington State, U.S.A. Canadian Journal of Fisheries and Aquatic Sciences 49:1246 -1256. Bilby, R.E., and P.A. Bisson. 1991. Allochthonous versus autochthonous organic matter contributions to the trophic support of fish populations in clear -cut and old- growth forested streams. Canadian Journal of Fisheries and Aauatic Science 49 :540 -551. Brett, J.R. 1956. Temperature tolerances in young Pacific salmon. Journal of the Fisheries Research Board of Canada 9:268 -323. Burton, T.A, G.W. Harvey, and M.L. McHenry. 1990. Protocols for assessment of dissolved oxygen, fine sediment and salmonid embryo survival in an artificial redd. Water Quality Monitoring Protocols Report Number 1. Idaho Division of Environmental Quality, Boise. Bustard, D.R., and D.W. Narver. 1975. Aspects of the winter ecology of juvenile coho salmon and steelhead trout. Journal of the Fisheries Research Board of Canada 32:667 -680. Chatters, J.C., V.L. Butler, M.J. Scott, D.M. Anderson, and D.A. Neitzel. 1995. A paleoscience approach to estimating the effects of climatic warming on salmonid fisheries of the Columbia River basin, p. 489 -496. In R.J. Beamish (editor). Climate change and northern fish populations. Canadian Fisheries and Aquatic Sciences Special Publication Number 121. FERC (Federal Energy Regulatory Commission). 1993. Draft Staff Report: Glines Canyon (FERC N0. 588) and El,"ha (FERC No. 2683) Hydroelectric projects, Washington. Volume 1. Office of Hydropower Licensing, Federal Energy Regulatory Commission, Washington, D.C. 41 Franklin, J.T., and C.T. Dyrness. 1984. Natural vegetation of Oregon and Washington. Oregon State University Press, Corvalis, Oregon. Fraser, F.J., E.A. Perry, and D.T. Lightly. 1983. Big Qualicum River salmon development project. Volume 1: a biological assessment 1959 -1972. Canadian Fisheries and Aquatic Sciences Technical Report 1189. Gorsline, J. 1992. Shadows of our ancestors. Readings in the history of Klallam -White relations. Empty Bowl Press, Port Townsend, Washington. Grette, G.B. 1985. The role of large organic debris in juvenile salmonid rearing habitat in small streams. M.S. Thesis, University of Washington, Seattle. Grier, N., E. Clough, and A. Clewell. 1994. Toxic Water. A report on the adverse effects of pesticides on Pacific coho salmon and the prevalence of pesticides in coho habitat. Unpublished Report. Northwest Coalition for Alternatives to Pesticides, Eugene, Oregon. Hartman, G.F., and J.C. Scrivener. 1990. Impacts of forestry practices on a coastal stream ecosystem, Carnation Creek, British Columbia. Canadian Bulletin of Fisheries and Aquatic Sciences Number 223. James River II, Inc. 1988. Response to May 28, 1987 request for additional information. Volume 3: Elwha Project (FERC No. 2683) and Glines Project (FERC No. 588). James River II Corporation. Johnson, P. 1995. Historical assessment of Elwha River fisheries. Unpublished Report, Olympic National Park, Port Angeles. Jones and Stokes Associates. 1991. Watershed characteristics and conditions inventory, Pysht River and Snow Creek watersheds. Unpublished Report, Washington Department of Natural Resources, Olympia. Larkin, P.A. 1977. Pacific salmon. In: J.A. Gulland (editor). Fish Population Dynamics. J. Wiley and Sons, New York. Lestelle, L.C., Rowse, M.L., and C. Weller. 1993. Evaluation of natural stock improvement measures for Hood Canal Coho Salmon. Technical Report #93 -1. Lestelle, L.C., and C. Weller. 1994. Summary Report: Hoko and Skokomish coho salmon indicator stock studies 1986 -1989. Technical Report #94 -1. Point No Point Treaty Council, Kingston, Washington. 42 Lichatowich, J.A. 1993a. The status of anadromous fish stocks in the streams of eastern Jefferson County, Washington. Unpublished Report, Jamestown S'Klallam Tribe, Blynn, Washington. Lichatowich, J.A. 1993b. Oral history of Dungeness River salmon. Part 1. Interview with Dick Goin. Unpublished Report, Jamestown S'Klallam Tribe, Blynn, Washington. Lisle, T.E. 1989. Sediment transport and resulting deposition in spawning gravels, North Coastal California. Water Resources Research 25:1303 -1319. Maser, C., R.F. Tarrant, H.M. Trappe, and J.F. Franklin. 1988. From the forest to the sea: A story of fallen trees. Pacific Northwest Research Station, U.S. Forest Service, General Technical Report PNW- GTR -229, Portland, Oregon. Mason, J.C., and D.W. Chapman. 1965. Significance of early emergence, environmental rearing capacity and behavioral ecology of juvenile coho salmon in stream channels. Journal of the Fisheries Research Board of Canada 22:173 -190. McHenry, M.L., J. Lichatowich, and R. Hagaman. 1996. Status of Pacific salmon and their habitats on the Olympic Peninsula, Washington. Unpublished Report, Lower Elwha Klallam Tribe, Port Angeles, Washington. McHenry, M.L., Morrill, D. and E. Currence. 1994. Spawning gravel quality, watershed characteristics and early life history survival of coho salmon and steelhead in five North Olympic Peninsula watersheds. Section 205J Report to the Washington Department of Ecology, Olympia, Washington. McHenry, M.L., S. Shaw, C. Toal, and J. Gorsline. 1995. Assessment of physical and biological conditions within. the Deep Creek watershed, North Olympic Peninsula, Washington, and recommendations for restoration. Report to the Washington Department of Natural Resources, Olympic Region, Forks, Washington. McHenry, M.L., E. Shott, G. Grette, and R. Conrad. In Preparation. Decadal changes in the large woody debris composition of Olympic Peninsula streams. Morrill, D.C., M.L. McHenry, R.J. Naiman, E. Currence. In Preparation. The effects of spawning gravel quality, inter gravel dissolved oxygen and streambed scour processes on coho salmon and steelhead survival'. 43 Naiman, R.J. and 8 others. 1992. Fundamental elements of ecologically healthy watersheds in the Pacific Northwest coastal ecoregion. p. 127 -188. In: R.J. Naiman (editor). Watershed Management: Balancing sustainability and environmental change. Springer - Verlag, New York. Narver, D.W. 1978. Ecology of juvenile coho salmon - Can we use present knowledge for stream enhancement? In: B.G. Sheppard and R.M.J. Ginetz (editors). Proceedings of the 1977 Northeast Pacific Chinook and Coho Salmon Workshop. Marine Fisheries Service (Canada) Technical Report 759. Nawa, R.K., C.A. Frissell, nd W.J. Liss. 1990. Life history and persistence of anadromous salmonid stocks in relation to stream habitats and watershed classification. Annual Progress Report. Oregon Department of Fish and Wildlife, Corvalis. Nehlsen, W., J.E. Williams, and J. Lichatowich. 1993. Pacific salmon at the crossroads: Stocks at risk from California, Oregon, Idaho and Washington. Fisheries 16:4 -21. Nickleson, T.E., J.D. Rogers, S.L. Johnson, and M.F. Solazzi. 1992. Seasonal changes in habitat use by juvenile coho salmon (Oncorhynchus kisutch) in Oregon coastal streams. Canadian Journal of Fisheries and Aquatic Science 49:783 -789. NRC (National Research Council). 1995. Upstream. Salmon and Society in the Pacific Northwest. Orsborn, J.F., and S.C. Ralph. 1994. An aquatic resource assessment of the Dungeness River system: Phase II. Physical channel analysis, hydrology and hydraulics. Unpublished Report, Jamestown S'Klallam Tribe, Blynn, Washington. Pait, A.S., A.E. DeSouza, and D. Farrow. 1992. Agricultural pesticide use in coastal areas: A national summary. Strategic Environmental Assessment Division, Office of Ocean Resources Conservation and Assessment, National Oceanic and Atmospheric Administration. Pearcy, W.G., and J.P. Fisher. 1986. Migration of coho salmon during their first summer in the ocean. Fishery Bulletin 2:173 -195. Per_tec. 1995. Hoko River basin watershed analysis: Resource assessment and watershed prescriptions. Washington Department of Natural Resources, Olympic Region, Fortis. Peterson, N.P., A. Hendry, and T.P Quinn. 1 °92. Assessment of cumulative effects on salmonid habitat: some suggested parameters and target conditions. Report TFW -F3 -92 -001, Center for Streamside Studies, University of Washington, 44 Peterson, N.P., A. Hendry, and T.P Quinn. 1992. Assessment of cumulative effects on salmonid habitat: some suggested parameters and target conditions. Report TFW -F3 -92 -001, Center for Streamside Studies, University of Washington, Seattle. Quinn, T. 1994. Factors affecting the freshwater and marine growth and survival of coho salmon from Big Beef Creek. Chapter IV. In: The effect of forest practices on fish populations. Final Report to the Washington Department of Natural Resources, Olympia. Reeves, G.H., F.H. Everest, and T.E. Nickelson. 1989. Identification of physical habitats limiting the production of coho salmon in western Oregon and Washington. General Technical Report PNW -GTR -245. U.S. Forest Service, Pacific Northwest Research Station, Portland. Salo, E.O. and W.H. Bayliff. 1958. Artificial and natural production of silver salmon (Oncorhynchus k1sutch) at Minter Creek, Washington. Washington Department of Fisheries Research Bulletin 4. Sandercock, F.J. 1991. Life History of Coho salmon. In: Groot, C., and L. Margolis, editors. Pacific Salmon Life Histories, University of British Columbia Press, Vancouver. Scarlett, W.J., and C.J. Cederholm. 1984. Juvenile coho salmon fall- winter utilization of two small tributaries of the Clearwater River, Jefferson County, Washington. In: J.M. Walton, and D.G. Houston (editors) Proceedings of the Olympic wildfish conference, Peninsula College, Port Angeles, Washington. Schuett- Hames, D., B. Conrad, A. Pleus, and K. Lautz. 1995. Salmonid spawning gravel scour module. Unpublished Report. Northwest Indian Fisheries Commission, Olympia, Washington. Seddell, J.R., and J.L. Froggart. 1984. Importance of streamside forests to large rivers: the isolation of the Williamete River from its floodplain by snagging and streamside forest removal. Internationale Vereinigung fur Theoretische and Angewandte Limnologie Verhandlungen 22:1828 -1834. Shaw, S. 1995. Mass wasting analysis of the Deep Creek drainage. Unpublished Report, Washington Department of Natural Resources, Olympic Region, Forks, Washington. Shaw, S. 1993. Geomorphic analysis of the North Fork Green Creek. .Unpublished Report, Washington Department of Natural Resources, Olympic Region, Forks, Washington. 45 Smith, R.D., R.C. Sidle, P.E. Porter, and J.R. Noel. 1993. Experimental removal of woody debris on the channel morphology of a forest gravel bed stream. Journal of Hydrology 152:153- 178. Tabor, R. W. 1987. Geology of Olympic National Park. Pacific Northwest National Parks and Forest Association, Seattle, Washington. Tabor, R.W., and W. J. Cady. 1978. Geologic map of the Olympic Peninsula, Washington. Miscellaneous Investigations Series, Map I -1994. United States Geological Survey, Tacoma, Washington. Tschaplinski, P.J. and G.F. Hartman. 1983. Winter distribution of juvenile coho salmon (Oncorhynchus kisutch) befroe and after logging in Carnation Creek, British Columbia and some implications for overwinter survival. Canadian Journal of Fisheries and Aquatic Science 40:452 -461. WDFW et al. 1994. Salmon and Steelhead Stock Inventory. Volume 2. Strait of Juan de Fuca Region. Washington Department of Fish and Wildlife, Olympia. Williams, R.W., R.M. Laramie, and J.J. Ames. 1975. A catalog of Washington streams and salmon utilization. Volume 1. Puget Sound Region. Washington Department of Fisheries, Olympia. Zillges, G. 1977. Methodology for determining Puget Sound coho escapement goals, escapement estimates, 1977 pre - season run size prediction and in- season run assessment. Technical Report 28. Washington Department of Fisheries, Olympia, Washington. 46