Between 1962 and 1990, Sacramento River winter-run chinook salmon were occasionally reared at Coleman National Fish Hatchery (NFH). In 1988, the Ten-Point Winter-Run Restoration Plan, which called for the artificial propagation of winter-run chinook salmon, was developed by NMFS, USFWS, CDFG, and U.S. Bureau of Reclamation (USBR) (NMFS 1988). The next year, Sacramento River winter-run chinook salmon were listed as an endangered species under the ESA. As part of an artificial propagation program intended to help avoid extinction and speed recovery, winter-run adults have been collected primarily at Red Bluff Diversion Dam (RKm 391) and Keswick Dam (RKm 486) in the mainstem Sacramento River and then transported to the Coleman NFH, where they are held until maturity. Attempts to hold winter-run adults in 1989 and 1990 at the Coleman NFH facilities were generally unsuccessful due to epizootic disease and fungal infections (Forbes 1992). The 1991 brood year effectively marked the beginning of the program. Changes in husbandry techniques and the construction of new holding facilities at the Coleman NFH greatly improved adult survival and spawning success in 1991 (Forbes 1992); however, the presence of infectious hemopoietic necrosis virus (IHNV), Ceratomyxa shasta, and other pathogens, may limit the effectiveness of the program.
Although releases of as many as 1.5 million winter-run chinook salmon smolts per year have been proposed, only about 100,000 fish have been released during the current recovery effort (NRC 1996). The primary constraint to increased production is the low number of adults available for spawning, as the broodstock collection permit for the program under the ESA allows for a maximum of 20 adults to be taken if less than 1,500 adults are expected to pass Red Bluff Dam (Forbes 1992). In January 1992, the first 11,582 juvenile winter-run chinook salmon that were reared at Coleman NFH were released directly into the upper Sacramento River. It was hoped that the fish would imprint on, and return to, their release site rather than to the Coleman NFH or Battle Creek, which has low flow and high temperature conditions during the time of the adult return migration. However, it appears that all of the adults recovered from these releases in 1995 returned to the hatchery site rather than the upper Sacramento River, which contains suitable natural spawning habitat (USFWS 1996b).
Winter-run adults at Keswick and Red Bluff Dams are selected according to return migration timing, and presumptive winter-run adults are further distinguished from spring-run fish by their spawning time. Natural variability in spawning time, in combination with the use of hormones to induce ovulation and spermiation, may result in the misclassification of fish. Based on DNA analysis, Hedgecock et al. (1995) concluded that several spring-run adults had been accidentally incorporated into the winter-run broodstock program.
In addition to the supplementation program, a portion of the juveniles derived from adults collected as broodstock are kept at the hatchery as part of a captive broodstock program, which provides for full-term rearing to the adult stage (Hedrick et al. 1995, Flagg et al. 1995a). The captive broodstock program was also initiated in 1991. The primary goals of the Sacramento River winter-run chinook salmon captive broodstock initiative are to provide a reserve of genetic material, should the natural run collapse, and to provide an additional source of eggs for the Coleman NFH program until conditions in the Sacramento River improve (CDFG 1995). To maximize future recovery options and to mitigate against the risk of mechanical failure, about 1,000 juveniles are transferred each year to the Bodega Bay Marine Laboratory (University of California at Davis) or the California Academy of Science's Steinhart Aquarium. The goal is for captive broodstock technology to provide about 200 mature adults per year to be spawned at Coleman NFH (CDFG 1995). Based on results obtained to date, adult growth, survival, and gamete quality appear to be lower under captive culture than in the anadromous program (USFWS 1996a).
The propagation of Sacramento River spring-run chinook salmon began in 1872 with the construction of the U.S. Fisheries Commission Baird NFH on the McCloud River, a tributary of the Sacramento River. Livingston Stone, the first manager of the station, noted that the spring run of chinook salmon on the Sacramento River were already "much depleted," and that artificial propagation efforts were needed to revitalize the fishery (Stone 1874). The Baird NFH collected eggs from returning spring- and fall-run chinook salmon. During the first decade of operation the majority of the eggs were shipped to the East Coast in an effort to establish runs there (Shebley 1922). Operations were suspended from 1884-1888 due to low numbers of returning adults. Although millions of eggs were collected, generally only one-quarter of the eggs were reared on site, with the surplus transferred to other stations--primarily the CDFG Mt. Shasta Hatchery (Shebley 1922). In 1902, the Baird NFH collected 7,375,520 eggs from the spring run; some two-thirds were transferred to the Eel River and the Mt. Shasta Hatchery (Titcomb 1905). Until 1911, it was hatchery policy to release chinook salmon shortly after yolk sac resorption (Clark 1929), and the success of these releases was probably limited. As a result of egg transfers, hatchery practices, and irrigation diversions on the Sacramento River, the spring run of chinook salmon returning to the McCloud River had dramatically dwindled by 1914 (Titcomb 1917, Clark 1929). During the 1920s, the spring run egg-take at the Baird NFH rarely exceeded one million eggs, and there were several years when no eggs were obtained (Leach 1924, 1928, 1932). The hatchery was abandoned in 1936 (Leach 1941), and the site was submerged under Lake Shasta following the completion of Shasta Dam in 1943.
In an effort resembling the GCFMP, from 1941 to 1946 chinook salmon attempting to migrate to areas above Keswick and Shasta Dams were trapped and transported to Deer Creek to spawn naturally (spring-run only) or to the Coleman NFH on Battle Creek for artificial propagation (Moffett 1949). The transportation program for spring-run chinook salmon to Deer Creek met with limited success (Moffett 1949). From 1943 to 1949 approximately 6,853,310 spring-run chinook salmon were released from the Coleman NFH (Cope and Slater 1957). Analysis of marked spring- and fall-run fish released from the hatchery suggested that 16% of the fish returning during the "spring run" (based on a September 25 cut-off date) were the progeny of fall-run parents, and 19% of the fish returning during the "fall run" were the progeny of spring-run parents (Cope and Slater 1957). Releases from the Coleman NFH ceased in 1953 (Appendix D). Following termination of the Coleman NFH spring-run chinook salmon program, there was no artificial propagation of spring-run chinook salmon until 1967 when the California Fish and Game hatchery on the Feather River began operation. The founding stock was derived from a run of fish returning to the Feather River. Since that time over 32 million spring-run chinook salmon have been propagated at the Feather River Hatchery, and about 80% of those have been released outside of the Feather River Basin (Appendix D). Furthermore, half of all spring-run releases for the entire Central Valley have been off-station and these fish may not show the homing fidelity of fish released from their home stream. Current release practices increase the potential for hatchery fish to interbreed with fish from naturally spawning populations.
The United States Fisheries Commission Baird NFH collected both spring- and fall-run chinook salmon for broodstock. Over the years of its operation, 1872-1936, the proportion of fall-run chinook salmon relative to fish from the spring run collected at the Baird NFH increased each year. Over the course of the next two decades, several other hatcheries were established on various tributaries of the upper Sacramento River, collectively taking as many as 100 million eggs annually from fall-run and late-fall run chinook salmon (Shebley 1922). In total, 317 million eggs (spring- and fall-run chinook salmon) were collected at the Baird NFH from 1872 to 1924, and 801 million eggs (fall-run chinook salmon) were collected at the Battle Creek and Mill Creek fish hatcheries from 1895 to 1924 (Clark 1929). Of these eggs, nearly 100 million were sent overseas and to the eastern seaboard of the U.S., and 61 million eggs and fry were sent to the Eel River (Clark 1929). Although large numbers of eggs were incubated during these early years, hatchery practices severely limited the survival of released fish (this was especially true from 1895 to 1910 when it was hatchery policy to release unfed fry) (Clark 1929).
In the San Joaquin River Basin, the artificial propagation of chinook salmon developed much later than in the Sacramento River. An experimental fall-run chinook salmon hatchery was located in Fresno County during the 1920s (Taft 1941); however, it was not until 1964 and 1971 that the Mokelumne and Merced Hatcheries began operations, respectively (NRC 1996). Most of the hatchery stocks of fall-run chinook salmon used in the San Joaquin River Basin have been imported from Sacramento River hatcheries (Appendix D).
From 1943 to 1946, fall-run chinook salmon attempting to migrate to areas above Keswick and Shasta Dams were trapped and transported to the Coleman NFH on Battle Creek for artificial propagation (Moffett 1949). Some 10,566 transported female fall-run chinook salmon were spawned at the Coleman NFH between 1943 and 1946 (Moffett 1949). Several thousand additional fall-run chinook salmon were left in the Sacramento River to spawn, or transported and released into Battle Creek (Moffett 1949).
From the late 1940s to the present, about 1.7 billion hatchery-produced fall-run and late-fall-run chinook salmon have been released into Central Valley streams (Table 6). Almost half of these were produced at Coleman National Fish Hatchery (which replaced the Battle Creek Hatchery station in 1944), the other half originated primarily from Feather River and Nimbus Hatcheries (NRC 1996). Since the early 1980s tens of millions of fall-run chinook salmon have been released into the extreme lower Sacramento River and in estuarine areas (NRC 1996) to avoid mortality associated with juvenile migration past irrigation diversions and other hazards.
Artificial propagation programs in the Central Valley have used primarily Sacramento River stocks; less than 1% of the fall-run chinook salmon released here have been from non-Sacramento River stocks. However, because of the large area occupied by this ESU, an intra-ESU transfer could involve transporting and releasing fish as far as 600 kilometers away from their hatchery of origin.
The artificial propagation of fall-run chinook salmon began in southern Oregon on the Rogue River in the late 1880s with hatcheries operated by canneries, most notably canneries owned by R.D. Hume (Cobb 1930, Kostow 1995). The U.S. Fisheries Commission began operating the Rogue River substation in 1900 as an egg collection and rearing site for spring-run chinook salmon (Titcomb 1904). Several million surplus eggs from the Rogue River substation were sent to a private hatchery at Wedderburn, Oregon on the Rogue River (Titcomb 1904). Additional egg collecting stations were operated intermittently during subsequent years in the Rogue River Basin on the Applegate River, Illinois River, Elk Creek, and Butte Creek. With the construction of the Oregon Game Commission Butte Falls Hatchery in 1916, salmon propagation on the Rogue River was increasingly dominated by state programs. By 1928, 85 million chinook salmon had been released into the Rogue River from state, federal, and private hatcheries (Cobb 1930).
Although the spring-run chinook salmon hatchery efforts in the Rogue River Basin did not begin in earnest until the mid 1970s, it is today one of the largest spring-run chinook salmon hatchery programs on the west coast of North America (Kostow 1995), with about 23 million hatchery-produced spring-run chinook salmon released into the Rogue River since the completion of the Cole Rivers Hatchery in 1974 (Appendix D). In 1993, nearly 1.5 million spring-run chinook salmon were released from the Cole Rivers Hatchery alone (Kostow 1995).
Compared to many of the other ESUs, the influence of fall-run chinook salmon artificial propagation in southern Oregon has been relatively minor. One exception, the Chetco River, has been stocked with almost 9 million fish since 1974, although these have been primarily of Chetco River stock (Appendix D). The other southern Oregon streams have received a total of about 5 million fall-run chinook salmon during the same period (Appendix D). The Rogue River, for example, is primarily a spring-run chinook salmon stream and not heavily stocked with fall-run chinook salmon; hatchery fall-run chinook salmon comprised only about 7% of the total adult run in 1987 (Cramer 1987).
Fall-run chinook salmon hatchery supplementation programs in some southern Oregon tributaries (Indian Creek, Rogue River Basin, Hunter Creek, and Pistol River) were intended to increase natural production; however, the results have been disappointing with a decrease in the effective population size for each river over the course of these programs (Kostow 1995). Furthermore, there has been an increase in the incidence of hatchery-derived strays between rivers in the region (Kostow 1995). Similar programs have been conducted in the Winchuck and Chetco Rivers, but hatchery-to-wild ratios are unknown in these rivers. The Winchuck River hatchery program was recently terminated. Hatchery fall-run fish released into Hunter Creek and the Pistol River are now being marked with coded-wire tags to more fully evaluate the impact of these programs (Kostow 1995). In December of 1992, the ODFW Coastal Chinook Salmon Management Plan was implemented to provide guidelines for stock transfers and to identify streams where stocking of hatchery fish should be excluded (Kostow 1995).
California coastal hatcheries and egg collecting stations began operating on several coastal streams in the early 1890s, but the first permanent facility was not established until 1910, with the construction of the Snow Mountain Station (currently known as Van Arsdale Fisheries Station) on the Eel River (Shebley 1922). Facilities on the Eel and Mad Rivers were constructed to rehabilitate depressed north coast populations (Kelly et al. 1990). A total of 95 million chinook salmon fry were released into California coastal rivers from 1875 to 1919, the majority (84 million) into the Eel River (Cobb 1930). Hatchery releases of fall-run chinook salmon since the 1970s have been relatively small, especially when compared to the large programs in the adjacent Sacramento River Basin (Appendix D). For example, the Smith River has received about 133,000 fall-run chinook salmon per year (NRC 1996), a fraction of the number of fish released into Sacramento River tributaries of similar size. The majority of the current coastal California fall-run chinook hatchery programs tend to use stock developed within basin, although these stocks may not be wholly native due to the long history of interbasin transfers that were common in earlier decades (CDNR 1931). The Russian River is a notable exception to this rule, having received artificially propagated fall-run chinook salmon from a variety of sources, most commonly Sacramento River stocks and the Great Lakes (which were stocked with a myriad of populations from Washington, Oregon, and California) (Appendix D). In the absence of existing permanent native runs of chinook salmon, local enhancement efforts south of San Francisco Bay in this area have generally used Sacramento River fall-run chinook salmon, although stocks from Washington, Oregon and the Great Lakes have been released there as well (NRC 1996). Spring-run chinook salmon artificial propagation has been very limited in the coastal river basins of California, with the exception of the Klamath River Basin (see ESU #5).
Early artificial propagation efforts in the Upper Klamath and Trinity Rivers began at the turn of the century. In 1896, over a million chinook salmon fry were introduced into the Klamath River from the Sacramento River (Snyder 1931). In 1890, a fish hatchery at Fort Gaston on Minor Creek, a tributary to the Trinity River, was established (Kirk 1994). During the operation of this hatchery (1890-98) eggs were collected from the Trinity and Sacramento (Baird NFH) Rivers and Redwood Creek, and the majority of the 2 million fry produced from this facility were released into the Trinity River and Redwood Creek (Snyder 1931). Several canneries near the mouth of the Klamath River also operated small hatcheries on an intermittent basis. The U.S. Fisheries Commission Hornbrook Hatchery (later known as the Klamathon Racks) on Cottonwood Creek (a tributary of the Klamath River) initially trapped rainbow trout and coho salmon, but in 1914 trapping operations were relocated on the Klamath River to intercept chinook salmon (Snyder 1931). On average, several million eggs were collected at this site annually. By 1916, nearly 17 million chinook salmon fry had been released into the Klamath River Basin (Cobb 1930). Surplus eggs were normally transferred to the CDFG hatchery at Sisson, California (later named the Mt. Shasta Hatchery) for incubation and rearing (Snyder 1931).
To mitigate the loss of spawning habitat caused by the construction of COPCO Dam (RKm 320) on the Klamath River in 1917, a CDFG hatchery was constructed on Fall Creek (RKm 316) and supplied with eggs from the Klamathon egg collection site (Shebley 1922). From 1916 to 1928, over 118 million chinook salmon eggs had been collected from the Klamath River (Snyder 1931). Although a substantial proportion of the fry and fingerlings produced from these eggs were returned to the Klamath River Basin, millions of eggs and fry were transferred to the Sacramento, Eel, and Mad Rivers (Shebley 1915 1922; Snyder 1931). The disposition of many millions of additional eggs is unclear. The Fall Creek Hatchery was closed in 1948, and although egg collections continued, no rearing facilities existed on the Klamath until 1966 (KRBFTF 1991).
The construction of Iron Gate Dam on the Klamath River (1962) resulted in the construction of the Iron Gate Hatchery (1965). Eggs for the Iron Gate Hatchery have primarily been collected from adults returning to the hatchery, although the hatchery has occasionally relied on spawners captured in the nearby Bogus Creek. Similarly the impact of the completion of the Lewiston Dam (RKm 249) on the Trinity River (1964) was mitigated by the construction of the Trinity River Hatchery (RKm 247) in 1963. Prior to the completion of the hatchery (1958-62), returning adult chinook salmon had been trapped downstream from the dam construction site, spawned, and their eggs incubated at Mt. Shasta Hatchery.
Iron Gate Hatchery has released primarily fall-run chinook salmon. Attempts to maintain a spring run from adults returning to the hatchery were intermittent and eventually abandoned. The Trinity River Hatchery has successfully maintained both fall and spring runs of chinook salmon. Both hatcheries have relied on returning adults to maintained their runs. Since 1965, the upper Klamath River has received about 7.3 million fall-run chinook salmon juveniles per year; almost all have been Klamath River stock (Appendix D). Since 1964, about 2.6 million fall-run chinook salmon and 1.5 million spring-run chinook salmon have been released in the Trinity River each year (Appendix D), all of which have been of Trinity or Klamath River origin.
Pathogens, specifically infectious hematopoietic necrosis virus (IHNV) and bacterial kidney disease (BKD), which are caused by Renibacterium salmoninarum, have been detected in juvenile and returning adult spring-run chinook salmon from the Trinity River Hatchery (PFMC 1994). These pathogens may have significantly limited the success of hatchery programs in the Klamath River Basin; for example, IHNV was associated with the loss of 20% of the spring-run chinook juveniles held at the Trinity River Hatchery (PFMC 1994). Another consequence of artificial propagation in this ESU has been the inadvertent hybridization of chinook and coho salmon at the Iron Gate Hatchery (Bartley et al. 1990). However, because this interspecies hybrid is sterile (Johnson 1988a), the long-term genetic effects of this hybridization are minimal while ecological effects would depend on the hybridization rate.
Artificial propagation efforts for chinook salmon in this ESU began in the late 1890s. By the early 1900s, there were hatcheries or egg-taking stations on most of the larger streams along the Oregon coast, especially the Yaquina, Alsea, Siuslaw, Umpqua, Coos, and Coquille Rivers (Cobb 1930, Wahle and Smith 1979). Before 1960, a substantial portion of the chinook salmon introduced into river basins in this ESU came from lower Columbia River (LCR) fall- and spring-run chinook salmon stocks--mostly from the Bonneville and Clackamas Hatcheries (Appendix D).
Chinook salmon populations in this ESU were considered to be mostly wild prior to 1960, based on the relatively low number of hatchery fish contributing to naturally spawning populations (Kaczynski and Palmisano 1993). However, the contribution of hatchery-reared fish relative to naturally spawning fish in this ESU has apparently increased since that time (ODFW 1995). Declining numbers of wild salmon prompted an increase in artificial propagation efforts. Improvements in hatchery rearing and release practices, feed formulation, and disease treatment have allowed hatcheries to produce fish that are larger, more fully-smolted, and healthier than fish produced before the mid-1960s (McGie 1980). Releases of larger smolts, in turn, have yielded a higher survival to adulthood than previous releases of fry and parr-stage fish (CBFWA 1990a). Furthermore, legislation enacted in the mid-1970s allowed the establishment of privately operated, for-profit hatcheries in Oregon (Wahle and Smith 1979). Private facilities operated in the Coos River and Yaquina River Basins until 1988 and 1989, respectively (NRC 1996). These salmon ranching operations released millions of smolts produced from spring- and fall-run broodstock primarily obtained from Oregon coastal rivers, such as the Rogue, Trask, and Yaquina (NRC 1996). In addition, a number of smaller cooperative hatcheries, built to restore depleted populations, are responsible for a substantial proportion of the current hatchery production (Appendix D).
Currently, most of the fall-run chinook salmon populations in this ESU are thought to have been minimally influenced by hatchery fish, which made up less than 10% of the spawning population (Kostow 1995). However, hatchery fish are thought to comprise up to 50% or more of the naturally spawning fish in the Salmon and Elk Rivers (ODFW 1995); Kaczynski and Palmisano (1993) estimated that 78% of natural spawners in the Elk River were of hatchery origin. Although fall-run chinook salmon hatchery programs are currently in operation in a number of basins, ODFW (1995) concluded "hatchery fish are not thought to be sustaining natural production," or "are not needed to sustain natural production" in most streams in this region. The influence of stray hatchery fish between basins may be significant; strays constituted some 20% of the "naturally spawning" run in the Sixes River (Kaczynski and Palmisano 1993).
Hatchery programs for spring-run chinook salmon have a significant impact on populations in the Trask and Umpqua River Basins. Hatchery contributions constituted between 40 and 60% of the total run in the North Umpqua River (ODFW 1995). Furthermore, the broodstock initially collected for the Rock Creek Hatchery (1955) on the North Fork Umpqua River may have been influenced by introductions of Rogue River spring-run chinook salmon in 1951. Low returns of adult spring-run chinook salmon over Winchester Dam (RKm 116) from 1946-48 (average, 2,404) prompted the release of 35,524 and 3,270 yearling spring-run chinook salmon from the Rogue and Imnaha Rivers, respectively (ODFW 1954). Although the number of fish released was small during this period, the hatchery fish released into the Rogue River contributed 20.9 and 12.6% of the total adult run in 1953 and 1954, respectively, due to their large size at release (ODFW 1954). In addition, the abundance of the fall-run chinook salmon in the North Fork Umpqua River increased from 12 in 1952 to 684 in 1955, largely related to introductions of fall-run chinook salmon from hatcheries on the Columbia River (ODFW 1954). Hatchery-derived spring-run chinook salmon in the Wilson, Nestucca, and South Umpqua Rivers are thought to now be abundant enough that they "may mask [abundance] trends in wild populations" (ODFW 1995).
Naturally produced fish account for the majority of chinook salmon in this ESU; however, in 1993, artificial propagation efforts were still substantial, with releases of 3,700,000 fall-run and 840,000 juvenile spring-run chinook salmon (Kostow 1995). Efforts by ODFW to utilize locally derived stocks in artificial propagation programs may reduce deleterious wild-hatchery fish interactions provided that local stocks have not been genetically altered by previous non-native introductions.
In response to declining numbers of chinook salmon in Grays Harbor drainages, the State of Washington constructed a hatchery on the lower Chehalis River in 1897. However, the facility was poorly sited and soon relocated to the Satsop River (WDFG 1902, Moore et al. 1960). In 1899, a hatchery (which still exists) was built on the Willapa River, and by 1917 additional hatcheries were operating on the Humptulips, North, and Naselle Rivers (WDFG 1920, 1921). On average, several million fall-run chinook salmon were released annually from state hatcheries from 1917 to 1941. The early years of artificial propagation in the Washington Coast ESU were marked by widespread importations of non-native stocks to fill hatcheries to capacity (WDFG 1916) due to the depressed size of local populations, primarily from overharvest (WDFG 1921). Initially, the Quinault National Fish Hatchery (1914) was operated primarily as a sockeye salmon facility (Titcomb 1917), although releases of chinook salmon increased steadily through the years. Most of the effort regarding artificial propagation in ESU 7 has focused on fall-run chinook salmon. Hatcheries on the Washington coast tend to be located near areas of commercial harvest, with two facilities in operation on the Quinault River, two on major tributaries entering Grays Harbor, and three on tributaries to Willapa Bay. In general, non-native fall-run chinook salmon stocks, primarily Green River hatchery-derived stocks, were used in ESU 7 watersheds prior to 1975. Since 1980 there has been a shift to the use of locally returning stocks (Appendix D).
Hatchery-reared spring-run chinook salmon have been released in only a few watersheds: the Sol Duc, Hoh, Quinault, and Wynoochee Rivers (NRC 1996). The impact of artificial propagation on spring-run chinook salmon populations has been modest, and with the exception of the Sol Duc River (which has received more than 9 million hatchery spring-run chinook salmon since 1972), no watershed has received more than 500,000 spring-run chinook salmon during the period covered by our database (Appendix D). The Sol Duc River spring-run chinook salmon stock was originally established from Cowlitz River x Umpqua River hybrids, with subsequent introductions of Dungeness River spring-run chinook salmon for a number of years between 1973 and 1988 (Appendix D). Although the Sol Duc River is managed for hatchery production only, it apparently has influenced nearby naturally spawning populations. In both the Sol Duc and Quillayute Rivers, similarities in run timing and a substantial incidence of natural spawning by stray Sol Duc Hatchery spring-run chinook salmon may have resulted in significant genetic exchange between the hatchery spring-run chinook salmon and natural summer-run chinook salmon populations (WDF et al. 1993). The draft scoping document for a proposed wild salmonid policy for the Washington Department of Fish and Wildlife (WDFW et al. 1994) explains the value of the Sol Duc River spring-run chinook salmon stock as follows (p. V-31):
There are a number of unique hatchery stocks that have developed over time, out of a variety of parent stocks. Spring-run chinook at the Sol Duc Hatchery, Deschutes River (Washington) chinook, several of the stocks at the Quinault National Fish Hatchery and others represent unique genetic units that deserve some protection in the same way that we want to maintain unique wild stocks as a resource for future needs.
In general, watersheds that enter the Strait of Juan de Fuca portion of this ESU have not been stocked with hatchery fall-run chinook salmon since 1981. However, the Hoko River, which was stocked with Puget Sound and Hood Canal fall-run chinook salmon stocks from 1950 through the mid-1970s, has been stocked since 1984 with juveniles produced from adults returning to the Hoko River and reared at the Makah NFH (Appendix D).
The impact of artificial propagation on coastal systems has not been fully evaluated. There appears to be some confusion regarding stock origin and the influence of hatchery fish in some populations in this ESU, especially in tributaries of Grays Harbor. For example, the current Humptulips River Hatchery stock of fall-run chinook salmon, which was derived from both wild spawners and hatchery returns (the hatchery was founded from a variety of local and non-ESU sources (WDF et al. 1993)) has been designated as being of "native" stock origin (Ashbrook and Fuss 1996), while naturally spawning fall-run chinook salmon in the Humptulips River have been designated as of "mixed" stock origin, due to mixing with non-local stocks (WDF et al. 1993), although no non-native fall-run chinook salmon have been introduced to the system since 1981 (Appendix D). In addition, a recent study of genetic stock diversity of Washington chinook salmon populations states: "All of the spawning populations in Grays Harbor [six were identified] are considered native chinook with few impacts from hatchery releases or releases from outside the basin" (Marshall et al. 1995, p. D-31). Another recent study, based in part on genetic diversity and life-history characteristics, determined that three of these six naturally spawning Grays Harbor populations were of mixed stock origin (WDF et al. 1993), suggesting that releases from outside the basin have had some impact on them. It appears that solid data regarding the influence of artificial propagation has not yet been compiled for at least some naturally spawning populations in this ESU.
The artificial propagation of chinook salmon in the Columbia River was quickly followed by the establishment of hatcheries on Puget Sound tributaries, with state-run facilities operating in the Nooksack, Skagit, and Samish River Basins before the end of the last century. James Crawford, then Commissioner of the Washington State Fish Commission (WSFC), wrote (Crawford 1894):
That the salmon industry is in great danger, by reason of the decrease in the supply of salmon, cannot be successfully denied, and unless some steps are immediately taken to repair by artificial propagation the ravages annually made by the different fishing appliances on our salmon supply, this industry ... will pass into history .
By 1902, eight state-run and two federally-run chinook salmon hatcheries were operating in this ESU, and new facilities were being constructed every few years (Moore et al. 1960). There are currently about 46 state, tribal, and federal facilities that regularly release chinook salmon juveniles into Puget Sound tributaries and over 50 cooperative state/public facilities that occasionally produce chinook salmon (Appendix D). Transfers of chinook salmon eggs to Puget Sound from other geographic regions, primarily the lower Columbia River, were commonplace in the early history of artificial propagation in this region. For example, by 1914, Columbia River chinook salmon had been released in many watersheds throughout Puget Sound. Increases in the commercial salmon catch subsequent to these stock transfers were assumed to be directly related to artificial propagation efforts: "The most convincing results are apparent from the practice of transplanting surplus eggs from one hatchery to another," and the increased abundance of Puget Sound chinook salmon at that time was seen as "the direct result of the transferring of the surplus chinook salmon egg take of the Columbia River to Puget Sound and other districts." (WDFG 1914, p. 17). The perceived benefits of inter-watershed stock transfers had a long-term impact on hatchery policies in Puget Sound and elsewhere. In 1924 state-operated hatcheries in Puget Sound collected 11,460,600 eggs from returning adults; however, an additional 6,000,000 eggs were transferred to Puget Sound from outside the region (Mayhall 1925). By 1928, almost 290 million chinook salmon fry, fingerlings, and yearlings had been released into Puget Sound tributaries (Cobb 1930). The emphasis on producing fish for harvest during the early part of this century resulted in widespread movements of chinook salmon between watersheds in this ESU (NRC 1996) (Appendix D). However, stock integrity and genetic diversity have recently become important management objectives as well, and policy revisions restricting some stock transfers have been initiated to reduce the impact of hatchery fish on natural populations (WDF 1991, WDF et al. 1993, Ashbrook and Fuss 1996).
The Green River fall-run chinook salmon stock has been the dominant hatchery stock in this ESU since the construction of the Green River Hatchery in 1907. Substantial numbers of Green River fish have long been released in many rivers, as well as numerous smaller watersheds and saltwater release sites throughout Puget Sound (Appendix D), raising concerns that this strategy may erode genetic diversity (Busack and Marshall 1995). Although reliance on this stock in hatchery programs is declining as a result of recent policy changes in inter-hatchery transfer of chinook salmon (WDF 1991), 20 hatcheries and 10 net-pen programs still regularly released Green River fall-run chinook salmon as late as 1995 (Marshall et al. 1995). In a recent determination of salmon genetic diversity units in Washington, Busack and Marshall (1995) reported: "The extensive use of this stock has undoubtedly had an impact on among-stock diversity within the South Puget Sound, Hood Canal, and Snohomish summer/fall GDU (GDU 17), but may also have impacted GDUs elsewhere in Puget Sound and the Strait of Juan de Fuca."
Chinook salmon abundance in watersheds throughout the Puget Sound ESU appears to be closely correlated with hatchery effort. The recent stock assessment by WDF et al. (1993) identified 28 fall- and spring-run chinook salmon stocks in Puget Sound from the Nooksack River to the Elwha River (boundaries of NMFS ESU 8). Seventeen of these 28 stocks were reported to be naturally produced runs, reflecting evidence that hatchery fish have had little or no influence on the spawning grounds. The status of 15 of the 17 (88%) natural Puget Sound chinook salmon stocks was classified as "critical," "depressed," or "unknown" (WDF et al. 1993). On the other hand, WDF et al. (1993) reported that 6 of the 28 Puget Sound chinook salmon stocks were of "mixed production," based on a conclusion that hatchery fish have made a significant contribution to the spawning population. All six hatchery-influenced stocks have been designated as "healthy." Therefore, there are several river systems in which a constant infusion of hatchery fish appears to have maintained population abundance to the point that the stocks have been determined to be healthy, albeit "mixed."8
In at least one case, artificial propagation appears to have benefitted a declining stock. Spring-run chinook salmon in the White River have experienced a tremendous decline in abundance since the turn of the century, due principally to pronounced habitat alterations, although the harvest rate has been and is still estimated to be over 60% (WDFW et al. 1996). Several artificial propagation programs were initiated in the 1970s to boost the abundance of stocks of spring-run chinook salmon. The most successful of these was the propagation of White River spring-run by culturing fish in net-pens through maturity or releasing juveniles from a remote hatchery site. As a result of these artificial propagation programs, as well as harvest reductions to protect returning adults, abundance of this stock has steadily increased to the point that the captive broodstock portion is currently being phased out, and the remote hatchery program will be phased out in the future (WDFW et al. 1996). On the other hand, spring-run chinook salmon recovery programs on the Nooksack, Skagit, and Dungeness Rivers have been terminated or dramatically curtailed because of diminishing returns or the potential for interbreeding between different hatchery stocks or between wild and hatchery fish (WDF et al. 1993).
The first hatcheries in the Columbia River Basin were constructed by private companies in response to the declining abundance of chinook salmon that followed habitat destruction and overharvest. The first hatchery on the Oregon side was constructed on the Clackamas River in 1876, and the first Washington hatchery was built on Baker's Bay near the mouth of the Columbia River in 1894 (Wahle and Smith 1979). The first state-operated hatchery in Washington, which was built in 1895 on the Lower Kalama River, is still in operation. In Oregon, several hatcheries were built around the turn of the century on the Clackamas River, although none of these is still in operation. The oldest operational hatchery on the Oregon side of the lower Columbia River was built in 1909 near the town of Bonneville (Wahle and Smith 1979). The first federal chinook salmon hatchery on the lower Columbia River was built on the Little White Salmon River in 1897 (Nelson and Bodle 1990). The first half of the twentieth century was marked by an explosive increase in hatcheries and hatchery production. For example, from 1913 to 1930, 319 million chinook salmon fry were released into the lower Columbia River by Washington State hatcheries alone (WDF 1934). Oregon state and federal hatchery efforts were on a similar scale. Federal hatcheries on the Big White Salmon and Little White Salmon Rivers collected 20-40 million eggs annually, and a large number of these were transferred to various Oregon and Washington state hatcheries. Although there were considerable cutbacks in the number of hatcheries during the Great Depression, egg production reported for Washington state hatcheries on the lower Columbia River from 1935 to 1939 was 143,000,000 (WDF 1936, 1937, 1938, 1939, 1940). After 1938, there was a dramatic increase in the number of chinook salmon hatcheries in the lower Columbia River, due primarily to federal obligations to mitigate harvest opportunities lost as result of the construction of upper Columbia and Snake River dams (Wahle and Smith 1979). There was an interruption in hatchery operations during World War II, when production declined to one-tenth of the prewar years at Washington State hatcheries. At present, about 25 ODFW, WDFW, and USFWS hatcheries release chinook salmon in this ESU. Since the 1960s, a large number of hatchery programs in the lower Columbia River have been dedicated to mitigating for lost production (Howell et al. 1985).
A variety of stocks were released from the early hatcheries, the majority being of lower Columbia River origin (Howell et al. 1985), although some upriver stocks were propagated as well (Appendix D). Presently, lower Columbia River fall-run chinook salmon hatchery stocks continue to make up the majority of all chinook salmon in ESU 9. A majority of spawners in Oregon tributaries to the Columbia River may be Big Creek Hatchery strays, based on CWT analysis, as well as Rogue River fall-run chinook salmon released in lower Columbia River streams (Kostow 1995). Since 1960, most natural fall run spawning on the Oregon side of the lower Columbia River has been attributed to hatchery strays (Olsen et al. 1992). In fact, straying, along with habitat degradation, overharvest, and competition from hatchery juveniles, has been identified as one of the major problems facing naturally spawning fall-run chinook salmon in Oregon's lower Columbia River tributaries (Kostow 1995). Oregon fall-run chinook salmon programs use a number of different broodstocks, including local and hatchery-origin "tule" stocks, and stocks imported from other areas. The Rogue River stock was introduced into several Columbia River tributaries to produce a south-migrating stock that would be available for harvest primarily by Oregon fishers (Appendix D) (Kostow 1995). About 70-75% of other lower Columbia River hatchery fall-run chinook salmon turn north and are harvested in Alaska, British Columbia, and Washington (Vreeland 1989).
Similarly, the fall-run chinook salmon populations in Washington tributaries are thought to be essentially one widely mixed stock as a result of straying and egg transfers between hatcheries (Howell et al. 1985, WDF et al. 1993, Marshall et al. 1995). The majority of natural spawners in the Grays, Elochoman, Cowlitz, Kalama, Washougal, and Klickitat Rivers have been of hatchery origin, and strays from several lower Columbia River hatcheries are often found in these streams (WDF et al. 1993, Marshall et al. 1995). Hatchery strays are also the most numerous spawners in several Washington streams not believed to originally have had a native run of fall-run chinook salmon, such as Abernathy, Germany, Mill, and Skamokowa Creeks (Marshall et al. 1995). Strays from Oregon's Rogue River fall-run chinook salmon program at Young's Bay have been observed in the Elochoman River and Abernathy Creek (WDF et al. 1993, Marshall et al. 1995). In 1982, upriver "bright" fall-run chinook salmon were released from the Little White Salmon NFH (WDF et al. 1993). The founding broodstock for various upriver "bright" stocks were collected by intercepting returning adults destined for Columbia River spawning sites above the Dalles Dam. Since the initiation of the upriver "bright" program at the Little White Salmon NFH, large numbers of upriver "bright" strays have been found naturally spawning in the Wind, White Salmon, and Klickitat Rivers (WDF et al. 1993). Similarly, in 1986 the Klickitat River Hatchery began releasing upriver "brights" in lieu of tule fall-run chinook salmon.
Spring-run chinook salmon populations in the lower Columbia River are all thought to be heavily influenced by hatchery programs. Approximately 1.5 and 10 million spring-run chinook salmon were released from Oregon and Washington hatcheries, respectively, in 1993. Populations of spring-run chinook salmon in the Sandy and Clackamas Rivers are considered by Oregon biologists to be a component of upper Willamette River hatchery populations due to many years of inter-hatchery transfer (Kostow 1995). Dam construction and volcanic episodes have eliminated most of the historic spawning habitat for spring-run chinook salmon on the Washington side of the lower Columbia River (Marshall et al. 1995). The Cowlitz River spring-run chinook salmon stock has received only limited transfers of non-native stocks, but is strongly influenced by hatchery-derived fish (WDF et al. 1993). Stocks on the Lewis and Kalama Rivers are a composite of the Cowlitz River spring-run chinook salmon stock and other lower Columbia and Willamette River spring-run chinook salmon stocks (WDF et al. 1993). Numerically, most of the spring-run chinook salmon spawning naturally in lower Columbia River tributaries on the Washington side are now hatchery strays (Marshall et al. 1995). All Washington populations of spring-run chinook salmon in the lower Columbia River are currently managed as populations of mixed origin (WDF et al. 1993).