This region includes the Oregon Coast, Washington Coast, and Puget Sound ESUs. Chinook salmon were abundant in this region near the turn of the century, when estimates based on peak cannery pack suggested peak runs of near one million fish in the three ESUs combined. This region includes the Coastal Range and Puget Lowlands ecoregions (see "Ecological Features") and is characterized by numerous short rivers and streams draining the coast ranges and west slope of the northern Cascade Mountains, with relatively few large rivers (Umpqua, Chehalis, and Skagit Rivers).
Chinook salmon in this region have been strongly affected by losses and alterations of freshwater habitats. Bottom et al. (1985) and Bishop and Morgan (1996) provide thorough reviews of habitat problems. Timber harvesting and associated road building occur throughout the region on federal, state, tribal and private lands. These activities may increase sedimentation and debris flows, reduce cover and shade, and may reduce recruitment of large woody debris to streams, resulting in aggradation, embedded spawning gravel, loss of pools, and increased water temperatures. Agriculture is also widespread in the lower portions of river basins and has resulted in widespread removal of riparian vegetation, rerouting of streams, degradation of streambanks, and summer water withdrawals. Urban development has substantially altered watershed hydrodynamics and affected stream channel structure in many parts of Puget Sound and the Oregon Coast.
This region (and parts of the southern coastal region discussed above) has experienced severe winter floods in recent years which could have affected chinook salmon habitat and survival of in-stream juveniles during the flood events. The following discussion summarizes information available regarding floods in February 1996.
Between November 1995 and April 1996, the Pacific Northwest and California experienced a series of storm and flood events. High winds, heavy rainfall, rapid snowmelt, numerous landslides and debris torrents, mobilization of large woody debris and high runoff occurred over portions of California, Oregon, Washington, Idaho, and Montana (USFS and USBLM 1996). These storms also had a potentially large effect on northern California and Oregon coast coho salmon and their freshwater habitats. Abnormally high rainfall and warm temperature, on top of already elevated stream levels and saturated soils resulted in the floods of February 1996; considered to be 100-year floods in many Oregon coastal basins (USFS and USBLM 1996, Bush et al. 1997). USFS and USBLM (1996) estimated landscape-scale habitat impacts from the February 1996 flood on federal lands in Washington and Oregon. They identified the Wilson-Trask-Nestucca, Siuslaw, and Alsea Basins as experiencing landslides, gullies/surface erosion, bedload deposition, channel migration, and LWD deposition, and considered the Wilson-Trask-Nestucca area as one of four areas with the highest rates of disturbance from the flood, and the Siuslaw as one of four areas with the second highest rates of disturbance from the flood. Pacific Watershed Associates (PWA undated) conducted aerial surveys to provide an assessment of the nature, magnitude and spatial distribution of watershed erosion and impacts to streams channels in the middle Coast Range, including the Smith (Umpqua), Siuslaw, Alsea, and Yaquina Basins. They report that areas with the greatest impact included Hadsall and Knowles Creeks (Siuslaw River) and Lobster Creek (Alsea River), and those watersheds with a combination of steep slopes, unstable bedrock geology, recent timber harvesting, and high road densities within an altitude range where precipitation intensities were probably the greatest (500 m. in the Coast Range). They also stressed that landslides were highly correlated with forestry management activities and originated from recent clear-cuts and forest roads at much higher frequencies than from wilderness or unmanaged areas. In addition to these observations, PWA concluded that the floods may have had long-term effects on watershed habitats. Siuslaw National Forest (SNF 1996) staff surveyed 500,000 hectares of central Oregon coast forests using aerial photographs to assess the frequency and character of landslides. They detected 1,686 slides, 41% of which were associated with roads, 36% with recent (<20 year old) clear cuts, and 23% with forested areas. They also found that subbasins in the southern portion of the area assessed (Coos, Umpqua, Siltcoos and Siuslaw) experienced from 1.5 to 2.5 times more landslides by area than more northern areas. They attribute this difference to both landtype associations of the basins and the differential intensity of the storm as it moved onshore. They also determined that "stabilized" roads (those treated to reduce failure) were less likely to be the source of large (>1700 m3) landslides than untreated roads.
With regard to impacts to in-stream coho salmon habitat, ODFW has conducted random resurveys of habitat for 105 reaches since the floods (Moore and Jones 1997). This survey effort indicated that along the North Oregon Coast (Salmon River to Columbia River), 7.5% of habitats received "no impact" (no perceivable impact), 60% of habitats received "low impact" (high water and scour and deposition impacts), 28% received "moderate impact" (channel modified impact), and 3.4% received "torrents" (and of these levels associated with debris torrents or dam break floods). Along the mid coast (Siuslaw River to Devils Lake tributaries), 2% of habitats received "no impact," 91% received "low impact," 7% "moderate impact," and 0.1% "torrents." Habitat changes included both positive and negative effects, depending on the area. Bush et al. (1997) noted that there were substantial changes in pool and riffle areas, large woody debris, and streambed substrates in streams following the floods, based on differences in stream reaches initially surveyed in 1992-95 and resurveyed in 1996. Decreases in pool area ranged from 10 to 50%, and largely resulted from a 60% loss of beaver pond habitat. Large woody debris decreased by approximately 25% from the initial surveys, although much of the lost wood had been pushed up onto the floodplain or out of the active channel. Overall, large amounts of gravel were added to most streams, and new gravel bars were common. Dewberry et al. (1996) documented changes in salmon habitats in Knowles Creek. (Siuslaw River). Twenty four debris torrents occurred in anadromous fish-bearing reaches of the basin, four of which exceeded 3,000 m2. Although the floods had little impact on parts of the basin, including an old-growth section, other areas were highly affected.
Within the last 50 years, over 2.5 billion spring-, summer-, and fall-run chinook salmon have been released from state, federal, and/or tribal hatcheries in this region, with the fall run constituting the majority of these releases. In addition, large, privately owned sea-ranching programs operated in recent years on the Oregon coast. A number of hatcheries already were in existence on rivers around Puget Sound by the turn of the century, and many of those are still in operation. In coastal areas, the earliest and most intense artificial propagation efforts have been, and continue to be, in coastal rivers near the mouth of the Columbia River. The majority of these hatcheries have been built primarily for fisheries enhancement, rather than mitigation for habitat loss. However, hatcheries on the Skagit, White, Skokomish, and Elwha Rivers operate to mitigate the loss of habitat due to dam construction (WDF et al. 1993, Kostow 1995). Although there have been numerous introductions of lower Columbia River chinook salmon stocks into the region, the majority of fish released have been derived from local stocks (Table 6, Appendix D). Some artificial propagation programs on the Oregon and Washington coasts have recently begun to alter their primary mission from fisheries enhancement to the supplementation of natural populations.
ODFW has identified 45 populations of chinook salmon in the range of this ESU (Kostow 1995). Historical abundance estimates for this ESU are available only from cannery pack data. Peak cannery pack was 30,967 cases in 1896, suggesting a peak run-size of about 225,000 fish at that time. Abundance at that time does not reflect "pristine" conditions, as extensive logging with associated splash dams were already impacting stream habitat.
Types of data available in this ESU were much the same as within the Oregon portion of the Northern California/Southern Oregon ESU. Punch card data and average estimated harvest rates were used to estimate recent spawning run-size from freshwater angler harvest. Survey data from spawner surveys conducted by ODFW were used to estimate trends in abundance. The only other data available that provided reasonably long time series were fish counts of spring and fall runs at Winchester Dam on the North Umpqua River.
The 5-year geometric mean of terminal run-size calculated from angler catch was approximately 170,000 fish (spawning escapement of 136,000) distributed among numerous spawning populations (Fig. 33, Appendix E). Most long-term trends in escapement indices were stable or increasing, with only one population declining at more than 10% per year; short-term trends were more variable, with a mix of increases and decreases (Fig. 34, Appendix E).
Bottom et al. (1985) cited low streamflows and high summer temperatures exacerbated by water withdrawals as problems for many streams (notably Tillamook Bay tributaries and Alsea, Siletz, Siuslaw, and Umpqua Rivers) and noted that agricultural and logging practices have led to serious riparian habitat losses. They also cited serious modification of stream structure by logging, splash dams, and widespread removal of beaver dams, but concluded that recent efforts have resulted in more stream miles being accessible to anadromous fish now than 100 years ago. Effects of recent floods were discussed for the Oregon and Washington Coastal Region above.
The first hatcheries were built in this area in 1902. Since the 1930s, artificial propagation programs have released nearly 400 million fall- and spring-run fish into this area, with nearly one-quarter of all the fish released coming from sources outside the ESU (Table 6, Appendix D). During much of this period, the impact of these releases may have been reduced by the large size of naturally spawning runs in most rivers. However, during the 1940s and 1950s many rivers were experiencing record low natural runs, and hatchery releases may have had a significant impact on local populations during this period (Kostow 1995). Chinook salmon from the Trask River have been used to establish hatchery broodstock in other systems in the Tillamook and Nestucca River Basins (Kostow 1995).
The contribution of hatchery-derived fish to total escapement is generally thought to be rather low (Kostow 1995). In 1990, the hatchery contribution to the Tillamook Bay fishery was only 15% (Kostow 1995). In contrast, hatchery contribution to total spawning escapement has been reported to be highest (approximately 50%) among fall-run chinook salmon populations in the Salmon and Elk Rivers (ODFW 1995). Additionally, hatchery-reared spring-run chinook salmon constituted 50% of the spring run on the North Fork Umpqua River in the 1980s, although currently the figure may be as low as 30% (Kostow 1995). Estimates of the impact of hatchery strays is limited, but in the Sixes River, hatchery strays were reported to constitute up to 20% of the natural spawners (Kaczynski and Palmisano 1993).
Freshwater/estuarine harvest rates are on the order of 20-25% (Nicholas and Hankin 1988). Ocean exploitation rates have ranged from 24% to 48%, with total exploitation rates in the range of 45-68%, and an average near 60% (brood years 1982-89) (PSC 1994).
Previous assessments of stocks within this ESU have identified several stocks at risk or of concern; however, the preponderance of stocks have been identified as healthy (Appendix E). Nehlsen et al. (1991) identified two stocks as at high extinction risk (South Umpqua River and Coquille River spring run), one stock as at moderate extinction risk (Yachats River fall run) and five stocks as of special concern. Of the 44 stocks within this ESU considered by Nickelson et al. (1992), 26 were identified as healthy (with 2 stocks containing small, variable runs), 2 as depressed (South Umpqua River and Coquille River spring-run chinook salmon), 7 as of special concern due to hatchery strays, and 9 of unknown status (4 of which they suggested may not be viable). Huntington et al. (1996) identified 18 stocks in their survey: 6 healthy Level I and 12 healthy Level II stocks.
Historical harvest of chinook salmon in this ESU reached a peak in 1911, when 26,490 cases were packed at canneries. This corresponds to a peak run-size of about 190,000 fish.
At the present time, run-size and spawning escapement in this ESU are monitored by WDFW and the Western Washington Treaty Indian Tribes. Management objectives, terminal fisheries and monitoring methods vary considerably over the ESU. Willapa Bay is managed for hatchery production and is monitored by WDFW (WDF et al. 1993). Since 1988, 65% or more of the natural escapement in Willapa Bay has consisted of hatchery fish (WDF et al. 1993). Escapement is monitored by redd counts, and natural production is not believed to be self-sustaining. Monitoring of Grays Harbor is also conducted by WDFW through redd counts. Most spawning populations in Grays Harbor are believed to have little hatchery influence.
In rivers further north, monitoring is conducted by the Western Washington Treaty Indian Tribes. Time series of spawning escapement estimates are relatively short, and the longest abundance data series are from tribal net fisheries conducted in the estuaries. Most spawning stocks are believed to be of native origin with little hatchery influence. Notable exceptions are Sol Duc River spring-run chinook salmon, which are an introduced stock, and the Quinault River fall-run chinook salmon stock, which is propagated as a Pacific Salmon Treaty indicator stock.
Recent average natural spawning escapement, the sum of 5-year geometric means for individual populations, has been over 50,000 spawners (Fig. 35, Appendix E). Long-term trends are about evenly split between increases and declines, but with most larger populations increasing (Fig. 36, Appendix E). Short-term trends are predominantly negative, strongly so in the Quillayute Basin and Willapa Bay tributaries.
All basins are affected (to varying degrees) by habitat degradation. Tributaries inside Olympic National Park have been least affected by human activities For other areas, major habitat problems are related primarily to forest practices, including mass wasting resulting in sedimentation in spawning grounds, lack of large woody debris, and lack of streamside shade. For example, WDF et al. (1993) reported that the Hoko River has been heavily impacted by past logging practices, with over 300 mass-wasting events recorded in the last 50 years. Clearing of instream wood was common practice until the 1970s, resulting in channel downcutting and bedload scour and fill which, in combination with moderate to high levels of fine sediments in gravel beds, affects egg survival in many areas. Bishop and Morgan (1996) identified a variety of critical habitat issues for streams in the range of this ESU, including changes in flow regime (Hoko and, Quillayute Rivers), sedimentation (Chehalis, Hoh, Hoko, and Quillayute Rivers), high temperatures (Chehalis, Hoko, and Quillayute Rivers), streambed instability (Hoko and Quillayute Rivers), estuarine loss (Chehalis River), loss of large woody debris (Hoko River), and loss of pool habitat (Hoko River). Of the streams they reviewed, only in the Queets and Quinault River Basins were chinook salmon not considered to be substantially limited by habitat problems. Upper basins of several streams in this region lie within Olympic National Park and are fully protected from effects of logging and most other habitat changes. The Puget Sound Salmon Stock Review Group (PSSSRG 1997) reviewed causes of declines in western Strait of Juan de Fuca and described habitat conditions for rivers in that portion of this ESU, concluding that timber harvest and hydromodifications have reduced both capacity and quality of salmon habitats.
WDF et al. (1993) classified 9 out of 31 stocks in this ESU as having cultured or composite production (indicating that a stock is sustained to some extent by artificial propagation). Some 319 million chinook salmon have been released into Washington coastal waters since 1952. Fall-run chinook salmon have been propagated in much larger numbers than spring-run chinook salmon (309 vs. 10 million). On average, approximately 19% of all hatchery releases have been from sources outside of the ESU. However, the Pysht, Hoko, and Chehalis Rivers have received proportionally larger introductions of fish from outside the ESU. Releases into these three rivers constitute more than half of the total of all non-ESU releases (Table 6, Appendix D).
Significant numbers of hatchery strays have been found in naturally spawning populations in the Satsop and Willapa Bay Rivers (Marshall et al. 1995), although their reproductive success is unknown. Furthermore, there has been considerable interbreeding between the non-native Sol Duc Hatchery spring-run chinook salmon stock and the native summer-run chinook salmon run in the Sol Duc River (WDF et al. 1993). With the exception of the Sol Duc Hatchery spring run, most of the introductions of non-native spring-run fish are thought to have been unsuccessful (WDF et al. 1993, Marshall et al. 1995).
Harvest rates on Washington coast chinook salmon stocks have been moderate, with ocean exploitation rates averaging 44-52%, and total exploitation rates averaging 48-56% (1982-89) for Hoko and Sooes stocks (PSC 1994).
Previous assessments of stocks within this ESU have identified several as being at risk or of concern, but more stocks have been identified as healthy than at risk (Appendix E). Nehlsen et al. (1991) identified one stock as extinct (Pysht River fall run), one as possibly extinct (Ozette River fall run), and one as at high risk of extinction (Wynoochee River spring run), although there is some question whether the Wynoochee River spring run ever existed (WDFW 1997a). WDF et al. (1993) considered 31 stocks within the ESU, of which 18 were reported to be of native origin and predominantly natural production. The status of these 18 stocks was 11 healthy, 4 depressed, and 3 unknown. The status of the remaining (not native/natural) stocks was nine healthy, two depressed, and two unknown. The Sol Duc River spring-run and Raft River fall-run chinook salmon were not considered an ESA issue by the BRT (stocks were not historically present in the watershed or current stocks are not representative of historical stocks) but was included to give a complete presentation of stocks identified by WDF et al. (1993). Huntington et al. (1996) identified 12 stocks in their survey: 1 healthy Level I stock (Quillayute/Bogachiel River fall run) and 11 healthy Level II stocks.
The peak recorded harvest landed in Puget Sound occurred in 1908, when 95,210 cases of canned chinook salmon were packed. This corresponds to a run-size of approximately 690,000 chinook salmon at a time when both ocean harvest and hatchery production were negligible. (This estimate, as with other historical estimates, needs to be viewed cautiously; Puget Sound cannery pack probably included a portion of fish landed at Puget Sound ports but originating in adjacent areas, and the estimates of exploitation rates used in run-size expansions are not based on precise data.) Recent mean spawning escapements totaling 71,000 correspond to a run entering Puget Sound of approximately 160,000 fish. Based on an exploitation rate of one-third in intercepting ocean fisheries, the recent average potential run-size would be 240,000 chinook salmon (PSC 1994).
Currently, escapement to rivers in Puget Sound and Hood Canal is monitored by WDFW and the Northwest tribes. Populations least affected by hatcheries are in the northern part of the sound in the Nooksack, Skagit, Stillaguamish, and Snohomish River systems.
The Nooksack River has spring/summer runs in the north and south forks. The North Fork escapement is monitored by carcass surveys and is influenced by a hatchery on Kendall Creek (part of a native stock rebuilding program). Escapement to the South Fork is monitored by redd counts, and the stock is believed to have little hatchery influence. Both stocks are considered critical by WDFW because of chronically low spawning escapements. The Skagit River supports three spring runs, two summer runs, and a fall run. Mean spawning escapement of the summer/fall run has been below the escapement goal and declining (Fig. 37-38, Appendix E). Terminal run-size has been declining, and escapement has been maintained at the expense of terminal fisheries. Of the five stocks identified by WDF et al. (1993), two are rated healthy, two depressed, and one of unknown status. On the Stillaguamish River, two runs have been identified. The combined escapement goal has been met only twice since 1978, and both runs are considered depressed. Of four runs identified in the Snohomish system, two are rated depressed, one unknown, and one as healthy. The single stock identified as "healthy" (Wallace River) is considered to be derived from hatchery strays and has experienced a severe recent decline.
The 5-year geometric mean of spawning escapement of natural chinook salmon runs in North Puget Sound for 1992-96 is approximately 13,000 (Fig. 37, Appendix E). Both long- and short-term trends for these runs were negative, with few exceptions. In south Puget Sound, spawning escapement of the natural runs has averaged 11,000 spawners (Fig. 37, Appendix E). In this area, both long- and short-term trends are predominantly positive.
In Hood Canal, summer/fall-run chinook salmon spawn in the Skokomish, Union, Tahuya, Duckabush, Dosewallips and Hamma Hamma Rivers. Because of transfers of hatchery fish, these spawning populations are considered a single stock (WDF et al. 1993). Fisheries in the area are managed primarily for hatchery production and secondarily for natural escapement; high harvest rates directed at hatchery stocks have resulted in failure to meet natural escapement goals in most years (USFWS 1997a). The 5-year geometric mean natural spawning escapement has been 1,100 (Fig. 37, Appendix E), with negative short- and long-term trends (except in the Dosewallips River).
The ESU also includes the Dungeness and Elwha Rivers, which have natural chinook salmon runs as well as hatcheries. The Dungeness River has a run of spring/summer-run chinook salmon with a 5-year geometric mean natural escapement of 105 fish (Fig. 37, Appendix E). The Elwha River has a 5-year geometric mean escapement of 1,800 fish (Fig. 37, Appendix E), but contains two hatcheries, both lacking adequate adult recovery facilities. Egg take at the hatcheries is augmented from natural spawners, and hatchery fish spawn in the wild. Consequently, hatchery and natural spawners are not considered discrete stocks (WDF et al. 1993). Both of these populations exhibit downward recent trends (Appendix E).
Habitat throughout the ESU has been blocked or degraded. In general, upper tributaries have been impacted by forest practices and lower tributaries and mainstem rivers have been impacted by agriculture and/or urbanization. Diking for flood control, draining and filling of freshwater and estuarine wetlands, and sedimentation due to forest practices and urban development are cited as problems throughout the ESU (WDF et al. 1993). Blockages by dams, water diversions, and shifts in flow regime due to hydroelectric development and flood control projects are major habitat problems in several basins. Bishop and Morgan (1996) identified a variety of critical habitat issues for streams in the range of this ESU including 1) changes in flow regime (all basins), 2) sedimentation (all basins), 3) high temperatures (Dungeness, Elwha, Green/Duwamish, Skagit, Snohomish, and Stillaguamish Rivers), 4) streambed instability (most basins), 5) estuarine loss (most basins), 6) loss of large woody debris (Elwha, Snohomish, and White Rivers), 7) loss of pool habitat (Nooksack, Snohomish, and Stillaguamish Rivers), and 8) blockage or passage problems associated with dams or other structures (Cedar, Elwha, Green/Duwamish, Snohomish, and White Rivers). The Puget Sound Salmon Stock Review Group (PSSSRG 1997) provided an extensive review of habitat conditions for several of the stocks in this ESU. It concluded that reductions in habitat capacity and quality have contributed to escapement problems for Puget Sound chinook salmon. It cited evidence of direct losses of tributary and mainstem habitat, due to dams; of slough and side-channel habitat, caused by diking, dredging, and hydromodification; and also cited reductions in habitat quality due to land management activities.
WDF et al. (1993) classified 11 out of 29 stocks in this ESU as being sustained, in part, through artificial propagation. Nearly 2 billion fish have been released into Puget Sound tributaries since the 1950s (Table 6, Appendix D). The vast majority of these have been derived from local returning fall-run adults. Returns to hatcheries have accounted for 57% of the total spawning escapement, although the hatchery contribution to spawner escapement is probably much higher than that, due to hatchery-derived strays on the spawning grounds. In the Stillaguamish River, summer-run chinook have been supplemented under a wild broodstock program for the last decade. In some years, returns from this program have comprised from 30% to 50% of the natural spawners, suggesting that the unaided stock is not able to maintain itself (NWIFC 1997a). Almost all of the releases into this ESU have come from stocks within this ESU, with the majority of within-ESU transfers coming from the Green River Hatchery or hatchery broodstocks that have been derived from Green River stock (Marshall et al. 1995). The electrophoretic similarity between Green River fall-run chinook salmon and several other fall-run stocks in Puget Sound (Marshall et al. 1995) suggests that there may have been a significant effect from some hatchery transplants. Overall, the pervasive use of Green River stock throughout much of the extensive hatchery network, that exists in this ESU, may reduce the genetic diversity and fitness of naturally spawning populations.
Harvest impacts on Puget Sound chinook salmon stocks have been quite high. Ocean exploitation rates on natural stocks average 56-59%; total exploitation rates average 68-83% (1982-89 brood years) (PSC 1994). Total exploitation rates on some stocks have exceeded 90% (PSC 1994).
Previous assessments of stocks within this ESU have identified several stocks as being at risk or of concern (Appendix E). Nehlsen et al. (1991) identified four stocks as extinct, four stocks as possibly extinct, six stocks as at high risk of extinction, one stock as at moderate risk (White River spring run), and 1 stock (Puyallup River fall run) as of special concern. WDF et al. (1993) considered 28 stocks within the ESU, of which 13 were considered to be of native origin and predominantly natural production. The status of these 13 stocks was: 2 healthy (Upper Skagit River summer run and Upper Sauk River spring run), 5 depressed, 2 critical (South-Fork Nooksack River spring/summer run and Dungeness River spring/summer run), and 4 unknown. The status of the remaining (composite production) stocks was eight healthy, two depressed, two critical, and three unknown. The Nooksack/Samish River fall run and Issaquah Creek summer/fall run were not considered an ESA issue by the BRT (stocks were not historically present in the watershed or current stocks are not representative of historical stocks) but were included to give a complete presentation of stocks identified by WDF et al. (1993).
The Lower Columbia River Region includes portions of the Coastal Range, Willamette Valley, and Cascades ecoregions (see "Ecological Features") and is characterized by numerous short- and medium-length rivers and streams draining the coast ranges and west slope of the Cascade Mountains, with a single large river (Willamette River). We have no estimates of historic abundance of chinook salmon in this region. Peak cannery pack for the entire Columbia River Basin occurred in 1883, when 629,400 cases were packed, suggesting a total run-size of about 4.6 million chinook salmon.
Chinook salmon in this region have been strongly affected by losses and alterations of freshwater habitats. Bottom et al. (1985), WDF et al. (1993), and Kostow (1995) provide reviews of habitat problems. Timber harvesting and associated road building occur throughout the region on federal, state, and private lands. These activities may increase sedimentation and debris flows and reduce cover and shade, resulting in aggradation, embedded spawning gravel, and increased water temperatures. Timber harvest in the Oregon portion of the region peaked in the 1930s, but habitat impacts remain (Kostow 1995). Agriculture is also widespread in the lower portions of river basins, and has resulted in widespread removal of riparian vegetation, rerouting of streams, degradation of streambanks, and summer water withdrawals. Urban development has had substantial impacts in the lower Willamette Valley, including channelization and diking of rivers, filling and draining of wetlands, removal of riparian vegetation, and pollution (Kostow 1995).
Intensive hatchery programs were initiated more than 100 years ago in this region. Nearly 4.5 billion hatchery-derived fish have been released during the last 70 years, equal to the total for all the other regions combined (Table 6, Appendix D). The majority of these have been "tule" fall-run chinook salmon released into the lower Columbia River for fisheries enhancement. Because of the advanced degree of maturation that "tules" exhibit at the time of freshwater entry, the economic value of these fish is rather low; therefore, efforts have been made to introduce Rogue River "bright" fall-run chinook and upper Columbia River upriver "bright" fall-run chinook into this region (WDF et al. 1993, Kostow 1995, Marshall et al. 1995). In addition, fall-run chinook salmon from the lower Columbia River were introduced into the upper Willamette River Basin beginning in the 1950s to exploit underutilized habitat.
We have no estimates of historic abundance for this ESU, but there is widespread agreement that natural production has been substantially reduced over the last century. Currently, spawning escapement to populations on the Washington side of the Columbia River are monitored primarily by peak fish counts in index areas (WDF et al. 1993). Peak index-area spawning counts are expanded to estimate total spawning escapement. In most lower Columbia River tributaries in Oregon, foot surveys are conducted and escapement estimates are based on peak spawner counts or redd counts (Theis and Melcher 1995), with dam counts available for the Sandy and Clackamas Rivers.
For fishery monitoring purposes, these individual spawning populations are combined into stock groupings: Lower Columbia River Wild, Lower Columbia River Hatchery, and Spring Creek Hatchery stocks of fall-run chinook salmon designated for fishery management purposes(WDFW and ODFW 1994, PFMC 1996b).
The ESU also includes spring-run chinook salmon in the Cowlitz, Lewis, Kalama, Sandy, and Clackamas Rivers. Estimates of spring runs to the mainstem Columbia River tributaries are routinely reported by fishery management agencies (WDFW and ODFW 1994, PFMC 1996b), with the exception of the spring run to the Clackamas River. For fishery monitoring purposes, the Clackamas River spring-run chinook salmon are included with the Willamette River. Cramer et al. (1996) reported escapement to the Clackamas River (as hatchery returns), North Fork Dam counts, and spawners below the dam (from Bennett 1994).
Recent abundance of spawners includes a 5-year geometric mean natural spawning escapement of 11,200 spring-run fish (1992-96) (Fig. 39, Appendix E). The fall run includes 29,000 natural spawners (Fig. 39, Appendix E) and 37,000 hatchery spawners (1991-95), but according to the accounting of PFMC (1996b), approximately 68% of the natural spawners are first-generation hatchery strays. Long-term trends in escapement for the fall run are mixed, with most larger stocks positive, while the spring run trends are positive or stable (Fig. 40, Appendix E). Short-term trends for both runs are more negative.
All basins are affected (to varying degrees) by habitat degradation. Major habitat problems are related primarily to blockages, forest practices, urbanization in the Portland and Vancouver areas, and agriculture in floodplains and low-gradient tributaries. Substantial chinook salmon spawning habitat has been blocked (or passage substantially impaired) in the Cowlitz (Mayfield Dam 1963, RKm 84), Lewis (Merwin Dam 1931, RKm 31), Clackamas (North Fork Dam 1958, RKm 50), Hood (Powerdale Dam 1929, RKm 7), and Sandy (Marmot Dam 1912, RKm 48; Bull Run River dams in the early 1900s) Rivers (WDF et al. 1993, Kostow 1995).
Hatchery programs to enhance chinook salmon fisheries in the lower Columbia River began in the 1870s, expanded rapidly, and have continued throughout this century. Although the majority of the stocks have come from within this ESU, over 200 million fish from outside the ESU have been released since 1930 (Table 6, Appendix D). A particular concern at the present time is straying by Rogue River fall-run chinook salmon, which are released into the lower Columbia River to augment harvest opportunities. Available evidence indicates a pervasive influence of hatchery fish on natural populations throughout this ESU, including both spring- and fall-run populations (Howell et al. 1985, Marshall et al. 1995). In addition, the exchange of eggs between hatcheries in this ESU has led to the extensive genetic homogenization of hatchery stocks (Utter et al. 1989).
Harvest rates on fall-run stocks are moderately high, with an average total exploitation rate of 65% (1982-89 brood years) (PSC 1994). The average ocean exploitation rate for this period was 46%, while the freshwater harvest rate on the fall run has averaged 20%, ranging from 30% in 1991 to 2.4% in 1994. Harvest rates are somewhat lower for spring-run stocks, with estimates for the Lewis River averaging 24% ocean and 50% total exploitation rates in 1982-89 (PSC 1994). Inriver fisheries harvest approximately 15% of the lower river hatchery stock, 29% of the lower river wild stock, and 58% of the Spring Creek hatchery stock (PFMC 1996b). The average inriver exploitation rate on the stock as a whole is 29% (1991-95).
Previous assessments of stocks within this ESU have identified several stocks as being at risk or of concern (Appendix E). Nehlsen et al. (1991) identified two stocks as extinct (Lewis River spring run and Wind River fall run), four stocks as possibly extinct, and four stocks as at high risk of extinction. The Sandy River spring run and Hood River spring and fall runs were not considered an ESA issue by the BRT (stocks were not historically present in the watershed or current stocks are not representative of historical stocks) but were included to give a complete presentation of stocks identified by Nehlsen et al. (1991). WDF et al. (1993) considered 20 stocks within the ESU, of which only 2 were considered to be of native origin and predominantly natural production (Lewis River and East Fork Lewis River fall runs). Nehlsen et al. considered the status of these two stocks to be healthy, and the status of the remaining (not native/natural) stocks as: 14 healthy and 4 depressed. Huntington et al. (1996) identified one healthy Level I stock in their survey (Lewis River fall run).
The spring run has been counted at Willamette Falls since 1946 (ODFW and WDFW 1995) but, counts were not differentiated into adults and jacks until 1952. In the first 5 years (1946-50), the geometric mean of the counts for adults and jacks combined was 31,000 fish. The most recent 5-year (1992-96) geometric mean escapement above Willamette Falls was 26,000 adults (Appendix E). Willamette River spring-run chinook salmon are targeted by commercial and recreational fisheries in the lower Willamette and Columbia Rivers. During the same 5-year period, the geometric mean of the run-size to the mouth of the Columbia River was 48,000 fish (PFMC 1997). The majority of the Willamette River fish are hatchery produced.
Estimates of the naturally produced run have been made only for the McKenzie River in 1994 and 1995 (Nicholas 1995). Nicholas (1995) estimated the escapement of naturally produced spring-run chinook salmon in the McKenzie River to be approximately 1,000 spawners. Primarily on the basis of professional judgement, they estimated the 1994-95 natural escapement of spring-run chinook salmon to the entire ESU as approximately 7,700 spawners, with 2,100 to 3,500 naturally produced natural spawners. However, Nicholas (1995) included the Sandy and Clackamas Rivers in their Willamette River spring-run chinook salmon unit; the BRT does not consider these introduced populations to be part of the ESU. Without these 2 rivers, the remaining escapement was approximately 3,900 natural spawners, with approximately 1,300 of these spawners naturally produced (Fig. 39, Appendix E). Long-term trends of escapement are mixed, ranging from slightly upward to moderately downward (Fig. 40, Appendix E). Short-term trends are all strongly downward.
Although the abundance of Willamette River spring-run chinook salmon has been relatively stable over the long term, and there is evidence some of natural production, it is apparent that at present production and harvest levels the natural population is not replacing itself. With natural production accounting for only one-third of the natural spawning escapement, it is questionable whether natural spawners would be capable of replacing themselves even in the absence of fisheries. Although hatchery programs in the Willamette River Basin have maintained broodlines that are relatively free of genetic influences from outside the basin, they may have homogenized the population structure within the ESU. Prolonged artificial propagation of the majority of the production from this ESU may also have had deleterious effects on the ability of Willamette River spring-run chinook salmon to reproduce successfully in the wild.
Habitat blockage and degradation are significant problems in this ESU. Available habitat has been reduced by construction of dams in the Santiam, McKenzie, and Middle Fork Willamette River Basins, and these dams have probably adversely affected remaining production via thermal effects. Agricultural development and urbanization are the main causes of serious habitat degradation throughout the basin (Bottom et al. 1985, Kostow 1995).
Historically, only spring-run fish were able to ascend Willamette Falls to access the upper Willamette River (Fulton 1968). Following improvements in the fish ladder at Willamette Falls, some 200 million fall-run chinook salmon have been introduced into this ESU since the 1950s. In contrast, the upper Willamette River has received relatively few introductions of non-native spring-run fish from outside this ESU (Table 6, Appendix D). Artificial propagation efforts have been undertaken by a limited number of large facilities (McKenzie, Marion Forks, South Santiam, and Willamette [Dexter] Fish Hatcheries). These hatcheries have exchanged millions of eggs from various populations in the upper Willamette River Basin. The result of these transfers has been the loss of local genetic diversity and the formation of a single breeding unit in the Willamette River Basin (Kostow 1995). Considerable numbers of hatchery spring-run strays have been recovered from natural spawning grounds, and an estimated two-thirds of natural spawners are of hatchery origin (Nicholas 1995). There is also evidence that introduced fall-run chinook salmon have successfully spawned in the upper Willamette River (Howell et al 1985). Whether hybridization has occurred between native spring-run and introduced fall-run fish is not known.
Total harvest rates on stocks in this ESU are moderately high with the average total harvest mortality rate estimated to be 72% in 1982-89, and a corresponding ocean exploitation rate of 24% (PSC 1994). This estimate does not fully account for escapement, and ODFW is in the process of revising harvest rate estimates for this stock; revised estimates may average 57% total harvest rate, with 16% ocean and 48% freshwater components (Kostow 1995). The inriver recreational harvest rate (Willamette River sport catch/estimated run size) for the period from 1991 through 1995 was 33% (data from PFMC 1996b).
The only previous assessment of risk to stocks in this ESU is that of Nehlsen et al. (1991), who identified the Willamette River spring-run chinook salmon as of special concern (Appendix E). They noted vulnerability to minor disturbances, insufficient information on population trend, and the special life-history characteristics of this stock as causes for concern.