A Historical Perspective on Salmonid Production from Pacific Rim ...

A Historical Perspective on Salmonid Production from Pacific Rim Hatcheries

Conrad Mahnkenl, Gregory Ruggerone2, William Waknitzl, and Thomas Flaggl

lNorthwest Fisheries Science Center, 2725 Montlake Boulevard East Seattle Washington, 98112-2097, U.S.A.

2Natural Resource Consultants, 4055 21st Avenue West Seattle, Washington, 98199, U.S.A.

Mahnken, C., G. Ruggerone, W. Waknitz, and T. Flagg. 1998. A historical perspective on salmonid production from Pacific rim hatcheries. N. Pac. Anadr. Fish Comm. Bull. No.1: 38-53

Annual hatchery production rates of chinook (Oncorhynchus tshawytscha), coho (0. kisutch) , sockeye (0. nerka) , pink (0. gorbuscha) , chum (0. keta), masu (0. masu} salmon, and steelhead trout (0. mykiss} were obtained from published and unpublished sources and compiled as a computer database. Pacific Rim hatchery production trends for the 40-year period from 1950-1992 were analyzed for all species from four geographic areas: Pacific Northwest (Washington, Oregon, Idaho, and California), Canada (British Columbia), Alaska, and Japan (Honshu and Hokkaido). Production of chum, sockeye, and pink salmon has increased dramatically in Japan, Canada, and Alaska in the past 20 years. Chinook, coho, steelhead, and masu have also experienced moderate increases in the same time period; however, production of coho, chinook and steelhead has declined since 1985. Trends in survival of hatchery fish over the period 1970-90 are demonstrated where data were available. We noted that survival of coho salmon was greatest for releases into large estuaries, such as southeast Alaska, Georgia Strait, and Puget Sound, than into drainages discharging directly into the Eastern Pacific Ocean. A negative cline in survival of coho salmon was observed moving both north and south from the center of the distribution of the species in British Columbia. Survival trends for fall chinook released north of Puget Sound tended to be the opposite of those released to the south. Survival of Japanese chum salmon released into the North Pacific Ocean has increased steadily from the mid-1960's to the present.

INTRODUCTION

Development of the North Pacific sahnonid hatchery system began in the late 19th century and has played a prominent role in enhancement of the salmonid resource in Pacific Rim nations since the 1950s. Until recently, the artificial propagation approach to enhancement of fisheries has not been seriously questioned, but the recent alarming declines in wild spawning stocks have forced a re-evaluation of industrial-scale hatchery production of north Pacific sahnonids. Declines have been observed in chinook, coho, and sockeye sahnon stocks in the Pacific Northwest; high harvest rates of wild fish in fisheries targeted on the more abundant hatchery stock., have continued, and high production of hatchery chum sahnon in Japan and both pinks and chum sahnon in

38

Alaska, declining fish size and altered return timing and age at maturity, have raised concerns over limits on ocean carrying capacity.

In the Pacific Northwest, recent Endangered Species Act listings of Redfish Lake sockeye in the Stanley Basin of Idaho, Sacramento River winter chinook in California, and fall, summer, and spring runs of chinook in the Snake River Basin of Idaho and Oregon have focused attention not only on habitat loss and fishery-related impacts on the wild stock" but on the genetic and demographic consequences of uncontrolled expansion of hatcheries as well (Meffe 1992; Nehlsen et al. 1991). Proposed recovery plans for listed Snake River chinooks and sockeye call for a limit on annual releases from Columbia Basin hatcheries to 1994 levels (Schmitten et al. 1995).

In Alaska, the hatchery successes that were hailed

in the mid 1980s are being assailed in the 1990s. The return of record numbers of hatchery pink and chum salmon and high abundance of natural fish in the North Pacific has led to record high catches and record low revenues to fishermen, and has brought forward new criticism of hatchery management strategies. Concerns over decreasing fish size in the hatchery-based fishery for chum salmon in northern Japan has led to a decision by the Japanese government to reduce hatchery releases. Similar concerns for declining size and increasing age at maturity observed in North Pacific stocks of five salmon species suggests that large-scale hatchery production is resulting in density-dependant growth reduction (Kaeriyama and Urawa 1992; Bigler et al. 1996).

This report provides managers with an historical data set of hatchery releases from Pacific Rim nations, excluding Russia, for the six Pacific salmon species and steelhead through 1992. In addition, an analysis of survival trends for hatchery coho and fall chinook from the eastern Pacific and chum from Japan are presented.

METHODS

Release Data

Annual production rates from Pacific Rim hatcheries were obtained for six species of Pacific salmon: chinook (Oncorhynchus tshawytscha), coho (0. ldsutch) , sockeye, (0. nerka) , pink (0. gorbuscha) , chum (0. keta) , masu (0. masu), and steelhead trout (0. myldss). We compiled this information from published and unpublished sources and organized it in a computer database. Production trends for the 40-year period from 1950-1990 were analyzed for all species from four geographic areas: Pacific Northwest (Washington, Oregon, Idaho, and California), Canada (British Columbia), Alaska, and Japan (Honshu and Hokkaido). Data from Russia and Korea were unavailable as a continuous historical series, however discontinuous data were available from three sources and were used to estimate total present Pacific Rim output (Heard 1995; Malmken et al. 1983; Konovalov 1980).

Release data from Alaska for chum, pink, coho, sockeye, and chinook salmon were obtained through Alaska Department of Fish and Game annual reports (McNair 1996). Alaska steelhead releases were supplied by Alaska Department of Fish and Game (Marianne McNair, Alaska Department of Fish and Game, P.O. Box 25526 Juneau Alaska, Pers. commun., Oct., 1996). Release data for all species from British Columbia were provided by Canadian Fisheries and Oceans (Ted Perry, Canadian Fisheries

39

and Oceans, 555 West Hastings Street, Vancouver, B.C., Pers. commun., Oct. 1996). Japanese release information on chum and pink salmon was supplied by the Hokkaido Salmon Hatchery (Masahide Kaeriyama and Shigihiko Urawa, Hokkaido Salmon Hatchery, 2-2 Nakano shima , Toyohira-ku, Sapporo 062, Japan, Pers. commun., Oct., 1996).

Until recently, a consolidated data set for hatchery production has been unavailable for the western United States (Wahle and Smith 1979; Malmken et al. 1983; Isaksson 1988). We compiled a comprehensive historical data set from previously unreported raw data forms archived by fishery agencies, annual reports of hatcheries, and electronic databases. We obtained earlier salmon release records from Oregon and Idaho and steelhead from Washington which were previously not available.

Survival Data

Survival information for coho and chinook salmon was obtained from coded wire tag (CWT) data bases maintained by the Pacific States Marine Fisheries Commission. The historical CWT data set was assembled by Maria Claribel Coronado-Hernandez (1995) and is contained in her doctoral dissertation. This seminal work synthesizes and analyzes, for the first time, spatial and temporal factors affecting survival of hatchery-reared chinook and coho salmon and steelhead in western North America over the time period for which coded wire tag recoveries are available (1971-1989). Release-recovery information is presented in the form of expanded CWT recoveries using only non-experimental production groups. Recoveries were standardized to the most common age at return using virtual population analysis (Coronado­Hernandez 1995).

We computed mean survival rates from this data set and used two-way analysis of variance to make statistical comparisons of mean survival values over time. We chose coho and fall chinook salmon for analysis of temporal and geographic variatIOn because of the large data sets available for these species. Pairwise comparisons between sampling dates were made using the Fishers PLSD test. All statements about statistical comparisons are based on the P < 0.05 significance level.

Survival data for Japanese chum salmon was provided by the Hokkaido Salmon Hatchery (Masahide Kaeriyama and Shigihiko Urawa, Hokkaido Salmon Hatchery, 2-2 Nakanoshima, Toyohira-ku, Sapporo 062, Japan, Pers. commun., Oct., 1996). These data were compiled through records of fishery catch and hatchery escapement to the Japanese coastal net fishery and to Hokkaido and Honshu hatcheries, respectively.

NPAFC Bulletin No.1

RESULTS - HISTORICAL PRODUCTION OF PACIFIC SALMON AT PACIFIC RIM

HATCHERIES

Coho Salmon

Coho salmon are among the most successful of hatchery-cultivated species in the Pacific Northwest and Canada. Coho salmon hatcheries were highly successful in returning adults to the fisheries in the 1970s, when record smolt-to-adult survivals were recorded in the British Columbia and Puget Sound regions.

Coho salmon production in western North America grew slowly from its inception at the tum of the century, and before 1940 the output from hatcheries never exceeded about 25 million fish annually (Fig. lA). The slow rate of growth can be attributed to failure of hatcheries to contribute to expansion of the fishery (McNeil and Bailey 1975). Following a period of reduced production during World War II, coho salmon hatcheries entered an industrial phase of expansion that lasted through the 1970s.

In the 1950s and 1960s, advances in the knowledge of feeds, diseases, and the early life history culture requirements of coho salmon led to improved post-release survival of hatchery fish (Lichatowich and McIntyre 1987). Coho salmon are released either as fry or as yearling smolts (25-30 g) and size at release had also been increased over the years and contributed to improved adult survival (Wahle and Smith 1979). These achievements were even more important in that they coincided with a period of rapid deterioration of freshwater habitat and blockage of major migratory pathways by hydroelectric dams. Given these successes, fishery managers came to believe that hatcheries were a means by which the Pacific Northwest could continue to develop its water resources for power, irrigation, and industrial or domestic use and at the same time, maintain fisheries at historic levels (Lichatowich and McIntyre 1987). Increased reliance on hatchery coho salmon led to the rapid expansion of production through the 1970s. In the late 1970s and early 1980s, private sea ranches added 9.3 million smolts per year to Oregon coastal production (Fig. lA) and helped lead to a record production of 198 million hatchery coho salmon in 1981. In the years that followed however, coho salmon production in the Pacific Northwest stabilized and began to decline (Fig. lA).

Begmning in 1989, a period of declining production began in most sectors of the coho salmon hatchery system. The decline in overall production from the contiguous western United States was partially offset by the added production of 40 million hatchery fish from Alaska and British Columbia. This

40

Mahnken et al. (1998)

decline can be attributed to various factors: Survival of adult hatchery fish declined following the oceanic regime shift in 1976 and a series of EI Nmos events in the late 1970s and early 1980s. These conditions resulted in reduced escapement, and hatcheries were unable to meet egg needs for full production. More restrictive inter-basin egg transfer policies were introduced to protect remaining wild populations, and reduced operating budgets at some hatcheries further reduced coho salmon production.

Chinook salmon

Like coho salmon, chinook salmon are released either as fry or as yearling smolts. Chinook salmon were the first salmon species to be artificially propagated in western north America and are artificially propagated in hatcheries along the eastern Pacific seaboard from California to Alaska. More chinook salmon have been produced from hatcheries than any other species in the Pacific Northwest. The first effort to artificially propagate chinook salmon in North America was at the Baird Fish Hatchery on the McCloud River in California in 1872. This hatchery was established to obtain chinook salmon eggs for transport to Atlantic Ocean tributaries to replace depleted Atlantic salmon (Salmo salar) runs. Today, the center of hatchery production is the Columbia River Basin where approximately 27 % of world chinook salmon is produced.

Fall chinook are the most commonly cultured life history type in both British Columbia and the Pacific Northwest. Also known as "ocean-type" (Healey 1991), fall chinook salmon most frequently inhabit coastal rivers, although ocean-type fish are cultured in the Columbia River as far upstream as the Methow River. Fall chinook salmon are sometimes released as fry but are more commonly reared for three months and released at hatcheries in the spring at approximately 7-10 g. To improve survival, some hatchery populations of fall chinook are reared to yearling size and generally released in the spring at 25-30 g as age-l smolts (Wahle and Smith 1979).

Spring and summer chinook salmon, or "stream­type" life history stocks (Healey 1991), are produced at hatcheries located primarily on large river systems of the Pacific Northwest, British Columbia, and Alaska. Spring chinook salmon are the predominant stocks produced in Alaskan hatcheries. Spring and summer stocks of chinook salmon are seldom released as underyearlings and are grown in hatcheries to the largest size, often over 100 g average, of any of the Pacific salmon species (NRC, 1995). Underyearling releases have been attempted at hatcheries where high rearing water temperatures result in accelerated growth, but survival to adult from releases of these intermediate sized fish is generally lower than those of

larger yearling fish (Zaugg et al. 1985, 1986).

Hatchery production of chinook sahnon began in Washington State in 1895 at the Kalama hatchery on the Columbia River, and production grew slowly to around 50 million fish released until the late 1930's (Fig. lB). Production dropped slightly during World War II, then accelerated following improvements in culture technology and construction of new hatcheries. The decade of 1950-1960 began the industrial phase of chinook sahnon hatchery production, as development in the Pacific Northwest resulted in loss of freshwater habitat. Growth of the fisheries also added pressure to the hatchery system to increase production. New hatcheries on the middle Columbia River were constructed to mitigate for lost habitat above Grand

Coulee dam and other mid-Columbia hydroelectric projects. Between 1960 and 1976, 30 hatcheries and 12 rearing ponds raised anadromous sahnonids in this region (Wahle and Smith 1979). Another surge in production occurred in Puget Sound hatcheries in Washington State during the 1960s. New lower Columbia River hatcheries and growing production in other sectors of the Pacific Northwest drove annual releases to more than 300 million chinook salmon smolts by the early 1980s. Production increases from British Columbia and Alaska hatcheries in the 1980s added another 100 million fish annually. By 1988, when production peaked, more than 420 million fry, fingerling, and smolts were being released from eastern Pacific hatcheries, a seven-fold increase from the base level of 59 million fish released in 1949.

Fig. 1 Hatchery production of (A) coho salmon and (B) chinook salmon juveniles from Pacific Northwest, British Columbia, and Alaska hatcheries, 1990-1992.

200~--------------------------------~--~

A COHO SALMON

150

U 100 G .. PACIFIC NORTHWEST

<D mJ BRITISH COLUMBIA V")

a II ALASKA

<D 50 <D

0:::: ..c

0 V")

U-500

<+-0 B CHINOOK SALMON V")

C 400 0

~ 300 [I .. PACIFIC NORTHWEST

200 II BRITISH COLUMBIA

• ALASKA

100

Release Year 41

NP AFe Bulletin No.1

Chinook salmon hatchery production began to decline in the late 1980s for the same reasons that coho salmon production is now falling.

Chum salmon

From the late 1880s until the recent expansion of chum salmon ranching in Japan, hatcheries released fry from rearing ponds to the stream soon after yolk sac absorption. Egg and fry development was accelerated in Hokkaido through the use of constant, 8°C ground water. These condition favored releases from early to mid-February when fry were sometimes subjected to severe conditions in the streams and coastal marine waters. With temperatures as low as 0-SoC during February and March, survival was minimal. In 1962, production-scale hatchery experiments were undertaken to delay release to a time of more favorable sea temperatures. These experiments involved feeding up to 300 million fry for short periods (Mayama 1985) and demonstrated that dry diets could increase body weight from 0.6 to 1.0 g in about one month. Fry fed on these diets could be released from Hokkaido hatcheries in May, when coastal water temperatures exceeded 10°C. These fry are usually released 50 d before average sea surface temperature reaches 15°C. Larger fish of 2-3 g average weight released . to coastal waters during spring periods of high primary and secondary productivity survived at a much higher rate than smaller ones. Adult returns from unfed fry released in 1950-60 averaged 1.2%; returns improved to 2.3% after 1966 as the percentage of fed fish increased (Isaksson 1988).

Japanese chum salmon catch was high prior to World War II, but dropped during the war years. Following the war, catch increased for a few years, then decreased as stocks were overexploited. Beginning in the 1970s, catch of Asian chum salmon rose again, primarily as a result of massive Japanese hatchery releases. From 1950 until 1970 Japanese hatchery production rose from 260 million fish released to around 580 million (Fig. 2A). Following improvements in adult contribution through the release of fed fry, and the loss of foreign fishing grounds to exclusive economic zones (EEZs), the Japanese chum salmon hatchery system entered a phase of rapid industrialization. Production rose from 260 million fry released in 1970 to 2 billion released in 1981. After the early 1980s, Japanese production of chum salmon leveled off.

The Japanese hatchery system is the largest in the world in terms of fish released and returned 78 million adult fish to the Japanese coastal fisheries in 1995 (H. Urawa, pers. commun.). The Japanese land­based fishery is almost entirely hatchery-dependant; wild stocks are virtually non-existent in northern

42

Mahnken et al. (1998)

Japan. Another interesting feature of the Japanese hatchery system is that it has reduced reliance on high­seas fisheries. The imposition of foreign EEZs has restricted the Japanese high-seas fishery on chum salmon destined for Russia and North America, but the loss to the national fishery has been more than replaced by hatcheries.

In North America, enhancement efforts for chum salmon also accelerated in the 1970s, primarily in Alaska and British Columbia (Fig. 2A), adding an additional 850 million fry to the already impressive releases by Japan for a total Pacific Rim production of nearly three billion chum salmon fry (exclusive of Russian hatchery releases).

Pink salmon

Pink salmon are second to chum salmon in the numbers of juveniles released into the North Pacific Ocean; accounting for 29% of the total reported in 1992 (Heard 1995). Pink salmon are released from most hatcheries at 0.5-2.0 g size (Isaksson 1988). Alaska is the largest producer of pink salmon and released more than 800 million juveniles in 1992 (Heard 1995; McNair 1996). Heard (1995) also notes that Russia is the second largest producer of hatchery pinks; with 584 million juveniles released in 1992.

Pink salmon releases remained low (less than 100 million) until the early 1980s, when an industrialization period began in Alaska, and numbers of released fish from North America increased tenfold to a total of 1.006 billion in 1992 (Fig. 2B). When added to Russian production, total world production of pink salmon was 1.590 billion produced in 1992. Pink salmon is the most recent of the Pacific salmon species to be industrially produced.

Sockeye salmon

Sockeye salmon were propagated in Alaska before the tum of the century to enhance existing runs, and millions of eggs were sent to the Atlantic coast in an attempt to establish runs there (Roppel 1982). However, initial culture efforts in Alaska were unsuccessful, and sockeye salmon programs were discontinued early in the 20th century (Allee 1990). Similarly, 11 sockeye salmon hatcheries built in British Columbia before 1917 produced no consistent benefits, and production ceased soon thereafter (Foerster 1968). In Washington, artificial propagation of sockeye salmon began in 1896 at the Baker Lake Station in the Skagit River basin and continued until this facility was closed in 1933. In addition to supplementing the run of sockeye salmon to Baker Lake, this facility was the source for the sockeye salmon introduced into Lake Washington where a strong run was eventually established (Kemmerich 1945).

Fig. 2 Hatchery production of (A) chum salmon and (B) pink salmon juveniles from Pacific Northwest, British Columbia, Alaska, and Japanese hatcheries, 1950-1993.

3000~------------------------------------~

v (!) V')

o (!)

(!) 0::::

..c V') .-

2500

2000

1500

1000

500

A CHUM SALMON

PACIFIC NORTHWEST

D BRITISH COLUMBIA

ALASKA

I2l JAPAN

UL 1250~--------------------------------------~ "+-

o B PINK SALMON

C 1000 o

750

500

250

o PACIFIC NORTHWEST

o BRITISH COLUMBIA

[] ALASKA

fa JAPAN

IJ') IJ') 0-.

o CO 0-.

IJ') CO 0-.

Release Year Sockeye salmon culture began in the Columbia

River in the 1940s at the Leavenworth Hatchery in eastern Washington state. This production effort attempted to mitigate for losses of sockeye due to construction of hydroelectric dams on the middle Columbia River (Mullan 1986). Smolts were

for a period of about 20 years, but disease and low returns forced abandonment of the program

the 1960s. Small smolt and fry-release programs still exist for sockeye in Puget Sound and on the Washington coast, but by and large, sockeye salmon culture in the Pacific Northwest is insignificant. Along with pink salmon, sockeye salmon production constitutes one of the smallest artificial propagation

43

programs for any species in the PacIfic Northwest. Recent attempts to reinvigorate sockeye salmon culture program in the Pacific Northwest include a combined hatchery and net-pen culture system operated by the State of Washington in Lake Wenatchee and a recovery program for the endangered Redfish Lake sockeye salmon in Idaho's Stanley Basin (Flagg et al. 1991, 1995).

British Columbia is now by far the largest producer of artificially propagated sockeye salmon, producing more than 290 million fry in 1993 (Fig. 3A). Fry are produced ill spawning channels containing gravel substrates, where returning adults spawn naturally. Fry are allowed to migrate

NPAFCBulletinNo.l

volitionally out of the channels upon swimup, usually into a lake. Spawning-channel culture of sockeye salmon in British Columbia constitutes the least invasive culture technique used to mass culture Pacific salmon. These extensive artificial propagation techniques require no feed, use natural spawning substrates, and require no handling of either juveniles or adults during rearing. Although spawning-channel culture of sockeye was initiated prior to the 1950s, the pro gram did not accelerate until the 1960s when output increased from 2 million fry in 1960 to 258 million in 1980 (Fig. 3A).

The Alaskan hatchery system is now the major producer of sockeye smolts in North America, and although a recent entrant (1974) into large-scale sockeye production, has grown rapidly and produced

Mahnken et al. (1998)

75 million smolts in 1994, or 21 % of total North American production (Fig. 3A). Alaska also releases age-O sockeye after brief holding in sea-pens.

Steelhead and masu salmon

Hatchery steelhead are produced entirely in North America (Fig. 3B), while masu salmon are only produced in significant numbers in Japan (Fig. 4). In terms of total production, steelhead and masu salmon are minor species, but in terms of their contribution to regional fisheries, steelhead programs are important and produced an estimated 738,000 adults annually from 1978 to 1987 (Light 1989). Like coho salmon and chinook salmon, steelhead coho salmon and chinook salmon, steelhead coho salmon and chinook

Fig. 3 Hatchery production of (A) sockeye salmon and (B) steelhead salmon juveniles from Pacific Northwest, British Columbia, and Alaska hatcheries and spawning channels, 1900·1992.

V')

C o

400

A SOCKEYE SALMON

300

[] PACIFIC NORTHWEST

200 fIl BRITISH COLUMBIA

• ALASKA

100

O+---~-'~~-=F=~~~~--~

40~----------------------------------------~

30

20

10

B STEElHEAD TROUT

[J PACIFIC NORTHWEST

1m BRITISH COLUMBIA

• ALASKA

O~~~~~~~~~~~~~~~~~~~~

8 ~

o ~

o ll') ~

Release Year 44

Fig. 4 Production of masu salmon juveniles from Japanese hatcheries, 1960-1992.

20

"'0 MASU SALMON <D V')

0 15 <D <D

0:::: ..c

V') lO-u: \.4-

0 V')

c 5 ,0

~ No Data

0 0 0 0 0 '-0 t-... <X) 0-0- 0- 0- 0-

Release Year

salmon, steelhead populations in Washington, Oregon, Idaho and California have experienced steep declines ill abundance in recent years.

Artificial propagation programs for steelhead first appeared in the late 1800s and are presently operating throughout the Pacific coast region from Alaska to California. Coastwide releases did not exceed 5 million fish for most of the first half of the century. Releases declined during the war years but increased rapidly from 1960 to a peak production of 35 million fish released in 1985-87. There was a steep reduction in steelhead production in Washington state (Fig. 3B) following the 1977 U.S. vs. Washington court decision that established tribal treaty fishing rights, but the state resumed former levels of production in 1984. Steelhead production dropped by about 20 % from 1989-1992. In 1992, more than 90% of total steelhead production was from hatcheries in the Pacific Northwest.

Unlike chum salmon in Japan, masu salmon production has actually decreased in the last two decades (Ohkuma and Nomura 1991). Masu salmon, which are highly dependent on the riverine environment for up to two years prior to outmigration to the sea and upon their early return as adults, have suffered due to degraded freshwater habitats. Channelization of stream banks with concrete and the construction of dams have damaged freshwater spawning and rearing habitat, and the practice of intercepting returning adults at the mouths of rivers for hatchery propagation has reduced the number of

45

naturally spawning adults (Ohkuma and Nomura 1991).

No dramatic rise in production releases have occurred in the years since 1960 for which data is available. Less than 10 million fish have been released annually (Fig. 4).

Total Pacific Rim Production

Total Pacific Rim production for all species combined, excluding Russia, is given in Figure 5. Four phases in the development of the Pacific Rim hatchery system are evident: 1) a developmental period beginning in the late 1800s and ending around 1970, during which a rudimentary hatchery husbandry was developed; 2) a period from 1970 to 1980, during which significant technological improvements in feed and disease control were made and new hatcheries were constructed; 3) an industrialization period from 1980 to 1990 when intense fishing, loss of freshwater habitat, and declining ocean productivity accentuated the construction of more mitigation and enhancement hatcheries and accelerated production; and 4) a post­industrialization hatchery period where survival declined, hatchery escapement goals for adult spawners were not reached, and reduced operating budgets resulted in a decline in production of chinook and coho salmon and steelhead in the Pacific Northwest and Canada while high production in the north Pacific for chum and pink salmon reduced north Pacific for chum and pink salmon reduced

Fig.5 Total hatchery production of coho, steelhead, masu, sockeye, chinook, pink, and chum salmon juveniles from the Pacific Northwest, British Columbia, Alaska, and Japan, 1950-1992.

5000.-------------------------------------~

ALL SPECIES "'0 <D VI

a <D <D

G:::

.c

.!!! LL 'I-

0 VI

C .0 .-~

4000

1000

o o lI) 0-.

1m

• D 0 [3 ..

lED II

COHO

STEELHEAD

MASU

SOCKEYE CHINOOK PINK CHUM

lI) lI) 0-.

lI) co 0-.

Release Year revenues to the fishennen and resulted in a leveling off or reduction of hatchery releases for chum and pink salmon. Total Pacific Rim production for all species in 1992 stood at more than 5.5 billion fry, fingerlings, and smolts released (Heard 1995).

RESULTS - TRENDS IN SURVIVAL OF HATCHERY FISH

Coho salmon

Regular CWT sampling of Alaskan coho salmon stocks began in the early 1970s. Short- and long-tenn trends in mean survival of hatchery coho salmon were evident for both southeast Alaska and the more northerly coastal regions. Survival of southeast Alaska hatchery stocks was much higher than that of fish from the coastal regions to the north. Southeast Alaska stocks showed two peaks in adult return: one in 1982 and another in 1988-89. Variance was highest during these peak years when mean survival exceeded 10%.

Mean survivals of hatchery coho salmon released mto tributaries of British Columbia's Strait of Georgia peaked in the early 1970s; these were the highest mean survIvals (> 25 %) recorded for any location within the North American range of the species (Fig. 6A). As in southeastern Alaska, variance was highest in years of high survival. Survival dropped from an average of around 20% in 1972-76 to less than 7% in the 1980s. The trend in survival since 1980 has

46

continued downward, reaching record lows of less than 2% in 1990. Fraser River hatchery stocks, which also enter the waters of Georgia Strait after release, showed a similar declining trend and comparable low survivals in the 19808.

A comparison of Puget Sound and Columbia River coho salmon stocks showed a similar trend of higher survival for those stocks released into coastal estuaries than those released into coastal oceanic environments (Fig. 6B). Puget Sound stocks exhibit consistently high mean survivals (7-13 %) from the early 1970s through the mid 1980s, with notable declines following the 1977 and 1983 EI Nmos. The decline associated with the 1983 EI Nmo continued at least through 1990. Again, variation was highest during years of highest survival. On the other hand, Columbia River hatchery stocks exhibited mean survivals of generally less than 5% from 1972-90. Like Puget Sound stocks, declines in survival of Columbia River stocks occurred after the EI Nmos of 1977 and 1983, but unlike Puget Sound stocks, the general downward trend was not evident after 1983.

We noted a striking difference in survival between hatchery coho salmon stocks released into the larger, protected coastal estuaries along the eastern Pacific seaboard (southeastern Alaska, Strait of Georgia, Puget Sound) and those released directly to the ocean. Figure 6A compares survival trends in a typical estuary, for example the Strait of Georgia, with those of all coastal stocks north of the Columbia River and south of the Columbia River. Although Georgia Strait

Fig. 6 Mean survival (coded-wire tags) of coho salmon released from Pacific coast hatcheries 1970-1990. (A) Georgia Strait estuary, coastal regions of north British Columbia/outer Vancouver IslandlWashington, and coastal regions of Oregon/California. (B) Puget Sound and the Columbia River. Error bars are 1 standard error. Asterisks denote differences (P<0.05) ANOVA, Fisher PlSD between successive means.

30.----------r------------------------~

20

0 10 > .-> ~

~ (/')

-+- 0 C 15 (l)

--0- PUGET SOUND

A OREGON/CALIFORNIA COAST

GEORGIA STRAIT

NO. B.C./VANC. IS./WASH. COAST

U B ~ . __ ...... _-

COLUMBIA RIVER (l) 0-

10

stock survivals fluctuate widely and exhibit a severe decline in the 1980s, they remained more than twice as productive (approximately 5 %) as the coastal stocks (approximately 2%) until 1987. After 1987, all stocks declined to survivals around 2 %.

A comparison of bi-decadal mean survival by region revealed another interesting feature of hatchery coho sahnon releases from 1970-90 (Fig. When arranged in order of declining latItudinal clines appeared among the coastal stocks, both north and south from the center of the geographic distributIon of the species, at about the latitude of Vancouver Island. This region had the highest survival (4%), with coastal California and coastal Alaska (at either end of the geographic having the lowest survivals, at 1-1. 5 %. Latitudinal anomolies were observed in large coastal estuaries such as the

Year

47

Strait of Georgia, Puget Sound, and southeast Alaska. These areas are known to be excellent areas for juvenile rearing and for survival of sahnonids (Healey, 1980; Simenstead et aI., 1982), with relatively high survivals of 5.5-7.5% during this period.

Fall chinook salmon

Fan chinook sahnon exhibit different survival characteristics from coho sahnon. Because of smaller size at release, surVIval of hatchery fall chinook sahnon is less than that of coho sahnon. Furthermore, survivals and trends in survival are distinctly different between fall chinook sahnon ._ .... _.J stocks north of Sound and those to the south 8). SurVIvals in Puger Sound, Strait of Georgia and outer Vancouver Island ill the

NPAFC Bulletin No.1 Mahnken et al. (1998)

Fig. 7 Mean survival of hatchery coho salmon by region, 1970-1990.

9,--------------,-------------------------, 8- .... • ... 7- T

0 • .L T T . > 6- • .L • 2: 5- 1

:::> T en • C 4- .L

0 3-<D ~ 2 Estuaries

0 Cl !:: ~ CI:: f:d Z LU

« V) ~ CI:: :::> « w 0 a:: CI:: I- -' > V) V) « CI:: :::> tti « LU 0 t;; (5 V)

<..') ;1i « u :::> CI:: CI:: Z 0.. 0 J: .... «

w I- > :::> <..') 0 CI:: W V) l-:::> 0

mid-1970s at between 3-4%, then declined sharply to less than 2 % in the 1980s following the EI Nifio of 1977 (Fig. 8A). A further decline to about 0.5% survival followed the 1983 EI Nifio and survival has continued downward since. Stocks south of Puget Sound showed the opposite trend, with survival rates of less than 1 % through 1982 rising to around 2% following the 1983 El Nifio (Fig. 8B). It appears that the 1983 El Nifio acted to enhance survival of southern fall chinook salmon in the mid-1980s rather than to decrease survival. However, mean survival of fall chinook salmon, aggregated by region, failed to show the same geographic cline as that of coho salmon.

Chum salmon

Survival of chum salmon released from Hokkaido hatcheries has followed a trend that appears to be less influenced by ocean conditions and more by improvements in hatchery technology (Figure 9). Survival of Hokkaido chum salmon has risen uniformly from 2.5 to 4% from 1965-1988, a period in which production increased from 550 to 970 million fry released. The number of juveniles released and their survival both increased with time. This apparent inverse density-dependant survival has been noted by McNeil (1991) who suggests that the relationship may be an artifact of improved hatchery technology, or

•

CI:: LU

~ CI::

« iii ~ :::> -' 0 U

48

Coastal

T • • 1 • ,. • ..l. •

I I

z 0 z :!S ~ 0 0 z <Xi V)

t3 <..') CI:: « 0 -' LU -' z « CI:: !!: « J: I- 0 -' V) -' V) « « « « -' U 3: 0 a t;;

U -' « -' « « 0 t;; « 0 U I- « V) U 0 « 0 u u

satiation of predators at higher levels of juvenile production. However, Kaeriyama (1996) presented evidence that other life history characteristics of enhanced chum salmon, namely decreased body size and an increase of age-at-maturity may indicate the beginning of a density-dependant effect of continued large-scale releases from Japanese hatcheries.

DISCUSSION

Researchers from North America and Japan have noted the dramatic decline in salmon stock abundance and body size in the southern portion of the species range in North America. These declines have been especially apparent over the past two decades (NeWsen et al. 1991; Bigler et al. 1996; Ricker 1981). However, while the abundance of stocks in the Pacific Northwest has declined, the abundance of populations to the north, both in Asia and North America, remain healthy and some have reached historical highs (Heard 1995; Kaeriyama and Urawa 1992; Burger and Wertheimer 1995; Zorpette 1995). Attempts have been made to assess the cause of these declines based on changes in freshwater conditions, fishing, or on variations in the marine environment (Beamish and Bouillon 1993; Cooper and Johnson 1992; Johnson 1984; Lawson 1993; Lichatowich 1993; Nickelson 1986; Northcote and Atagi 1994; Pearcy 1992; Richards and Olsen 1993; Olsen and Richards 1994;

Fig.8 Mean survival (coded-wire tags) of fall chinook salmon released from Pacific coast hatcheries 1970-1990. (A) Puget Sound, Strait of Georgia, and outer Vancouver Island regions. (B) upper and lower Columbia River, and Washington and Oregon coastal regions. Error bars are omitted for simplicity. Asterisks denote differences (P<0.05) ANOVA, Fisher PLSD between successive means.

5

4

3

2

0 > .-> I.-::)

V) 0 -+-C

3

Q) 2.5 U

I.-

Q) 2 CL

1.5

0.5

A

B ----.-----0--

-a---+---

----0---- PUGET SOUND

-0- GEORGIA STRAIT

_.-.... _- OUTER VANCOUVER IS.

UPPER COLUMBIA RIVER OREGON COASTAL WASHINGTON COASTAL LOWER COLUMBIA R.

o CO 0-..

Year

l.{') CO 0-..

o 0-.. 0-..

Fig.9 Mean survival of chum salmon released from Japanese hatcheries in Hokkaido, 1963-1988.

5

0 4 .2: 2:

3 ::J C/) -C ()) 2 U ~

()) D-

O (") '0 0-..

HOKKAIDO CHUM SALMON

Brood Year 49

CO CO 0-..

NP AFe Bulletin No.1

Francis and Sibley 1991; Kaeriyama 1996). Most researchers have concluded that, whatever the major factor(s) affecting survival of Pacific salmonids, they are most likely to occur in the ocean environment.

Cooper and Johnson (1992), compared trends in abundance of Washington, Oregon, and British Columbia steelhead and concluded that there were similarities in trends over the entire geographic range that indicated common factors were responsible for the observed changes in survival. Because freshwater, estuarine, and nearshore conditions differ considerably from year to year within this region, they concluded that these factors alone could not explain the similarities in steelhead survivals. They suggested that similarities in steelhead abundance trends in widely separated geographical regions indicated that common factors were responsible for the observed declines, and that oceanic conditions were responsible. Olsen and Richards (1994) came to a similar conclusion while working with aggregated coastwide chinook salmon production data, namely that similar chinook salmon run-size trends can be observed between several west coast river basins, and that the data support the hypothesis that ocean conditions have had a marked and uniform impact on chinook salmon production in the Pacific Northwest. Lichatowich (1993) has pointed out that the magnitude of oceanic environmental changes and their impacts on salmon survival may be so large as to mask changes that occur in the freshwater habitat. He cautioned that this may cause managers to falsely attribute increased ocean survival to restoration effects in freshwater. Hilborn et al. (1993) further emphasizes the same point by stating that attempts to understand the impact of in-river (Columbia River) actions on survival will be confounded by changes in ocean conditions.

Coded-wire tag data shows that for the period 1970-1990, coho salmon adult survival was highest for stocks released into large coastal estuaries. Survival in these estuaries is typified by widely fluctuating mean survivals. Conversely, survival of hatchery coho salmon released into coastal regions that lack protective coastal estuaries is typified by lower, more constant survival. However, differences in survival between estuarine and coastal releases of fall chinook salmon are not as dramatic, with regions like outer Vancouver Island and coastal Oregon performing as well or better than Puget Sound and the Strait of Georgia. It is possible that such factors as size and orne of entry to seawater, location and length of time in estuaries prior to outmigration, and predation may influence differences in absolute survival and temporal trends between the species.

Coho salmon survivals were depressed following the unusually strong El Nino events of the past two decades (Fig. 6) and continued to decline throughout the 1980s. Depressed survival associated with El

50

Mahnken et al. (1998)

Nino events is most evident in regions where survival has been historically high (southeast Alaska and Georgia Strait following the 1982-83 EI Nino, Puget Sound following the 1976-77 EI Nino). In the past 40 years, nine EI Ninos have affected the coastal regions of the eastern Pacific. In the 1970s and 1980s, the coastal regions of the Pacific Northwest and Canada have been beset by a series of four moderate-to-strong El Nino events, most notably the 1982-83 EI Nino, which by many measures was the strongest this century. Since 1970, El Ninos have occurred in 1972-73, 1976-77, 1982-83, and 1987-88.

A strong negative cline in survival of coho salmon is observed moving both north and south from the center of the species distribution in British Columbia, with stocks in western Alaska and California exhibiting the lowest survival. Fall chinook salmon, on the other hand, do not show the same strong latitudinal cline.

Fall chinook salmon survival, although apparently also affected by EI Nino events, seems directed by other external factors. Stock survival north of the Columbia River peaked in the mid 1970s, while survival for regions south of the river peaked in the mid-1980s. Furthermore, mean survival of fall chinook salmon aggregated by region failed to show the same geographic cline as coho salmon. The Strait of Georgia estuary produced highest survivals, but Puget Sound fall chinook salmon did not produce higher survivals than Coastal Oregon, and produced only slightly better survival than outer Vancouver Island or coastal California. This is surprising, given the well-documented importance of estuaries for growth and survival of juvenile chinook salmon (Healey 1991; McCabe et al. 1986). Nevertheless, there is some indication that hatchery fall chinook salmon juveniles spend less time in estuaries than wild juveniles which may reduce the benefits of such areas to artificially propagated fish (Levings et al. 1986). It may be that the overall lower survival of fall chinook salmon masks regional geographic differences so evident with coho salmon.

It is tempting to postulate a cause and effect relationship between the occurrence of EI Nino events and declines in survival of hatchery fish in the eastern Pacific, but no convincing ecological relationship exists. Climate conditions are known to have changed recently in the Pacific Northwest. Most Pacific salmonid stocks south of British Columbia have been affected by changes in ocean production that occurred during the 1970s. Pearcy (1992) and Lawson (1993) attribute this decline largely to ocean factors, but do not identify specific effects. However, given the increased frequency of EI Nino events in the past two decades, and large-scale secular warming of the region (Freeland 1990), it is certainly plausible that there is at least some response to EI Nino events in the form

of reduced survival of the species. Survival of coho aud chiuook sahuon iu the

southern regions may be driven by au entirely different set of regional oceau/climate conditions thau those governiug survival of chum, piuk, aud sockeye salmon; the most abundaut species iu the northerly regions. Two forms of decliue were evident iu coho salmon survival curves: sharp, short-term decliues associated with El Nifio events, aud a more prolonged long-term decliue, typical of the late 1980s, that may be related to general warmiug trends iu Pacific waters along the eastern Pacific seaboard (Freelaud 1990; Welch et al. 1995; Beamish aud Bouillon 1993). Stocks released from southern regions are believed to migrate in a narrow coastal corridor that is characterized by highly variable iuterannual chauges iu flow, temperature, aud current (Pearcy, 1992). Those species aud stocks released iuto the more northerly regions enter a much larger area of acceptable oceau conditions that is greatly iufluenced by the Aleutiau low pressure system (Beamish aud Bouillon 1993).

Hatcheries have played a major role iu supplyiug salmon aud trout to the common property fishery iu the Pacific Northwest. But with the near catastrophic decline iu the population of southern stocks aud overabundauce of the northern stocks, we have entered a new era iu the operation of hatcheries that cannot help but impact the traditional users of hatchery fish. The two-pronged dilemma is, how cau we iustitute hatchery reform to both protect aud recover wild/natural stocks aud at the same time maintaiu some reasonable level of harvest for the fishers?

The present hatchery system was developed to produce au iucreasiug supply of hatchery fish iuto a constaut oceauic ecosystem believed to be nearly limitless iu its capacity to accommodate juvenile sahuonids. Until recently, these oceauic ecosystems were believed to be stable, iuternally regulated, aud to behave iu a determiuistic manner. The more current view is of au open system iu near cons taut flux--a system without long-term stability aud one that is often under the iufluence of stochastic factors, mauy origiuating outside the ecosystem itself (Maugel et al. 1996). The modem view of the ecosystem is one characterized by ecological uncertaiuty. Lack of understaudiug of this principle aud its impact on interannual variability of sahuonid abundauce aud survival has acted agaiust wild fish populations through unregulated propagation aud harvests of hatchery fish. Perhaps it is time for fishery mauagers to regulate hatchery releases to accomodate decadel­scale variation iu oceau productivity.

51

REFERENCES

Allee, B. 1990. Sockeye sahuon enhaucement iu Alaska. 1990 Sockeye culture workshop, Anchorage, Alaska, July 10-11, 1990. p. 1.

Beamish, R., aud D. Bouillon. 1993. Pacific sahuon production trends iu relation to climate. Cau. J. Fish. Aquat. Sci. 50: 1002-1016.

Bigler, B.S., D.W. Welch, aud J.H. Helle. 1996. Decreasiug size of North Pacific sahuon (Oncorhynchus spp.): possible causes aud consequences. Cau. J. Fish. Aquat. Sci. 53: 455-465.

Burger, e.V., aud A.e. Wertheimer. 1995. Pacific sahuon iu Alaska. In E.T. LaRoe, G.S. Farris, C.E. Puckett, P.D. Dorau, aud M.J. Mac (eds.), Our liviug resources: a report to the nation on the distribution, abundauce, aud health of U.S. plants, auimals, aud ecosystems. U.S. Dept. of Interior, NBS, Washiugton, D.C., pp. 343-347.

Cooper, R., aud T. Johuson. 1992. Trends iu steelhead abundauce iu Washiugton aud along the Pacific Coast of North America. Wash. State Dep. Wildl, Fish. Mauage. Div., Rep. 92-20.

Coronado-Hernaudez, M.C. 1995. Spatial aud temporal factors affectiug survival of hatchery­reared chiuook, coho, aud steelhead. Ph.D. Thesis, Univ. Washiugton, Seattle, 235p.

Flagg, T.A., C.V.W. Mahuken, aud K.A. Johuson. 1995. Captive broodstocks for recovery of Snake River sockeye sahuon. Am. Fish. Soc. Symp. 15:81-90.

Flagg, T.A., J.L. Mighell, T.E. Ruehle, L.W. Harrell, aud C.V.W. Mahuken. 1991. Cle Elum Lake restoration feasibility study: fish husbaudry research, 1989-1991. Report to Bonneville Power Admiuistration, Contract DE-AI79-86BP64840, 52 p. (Available from Northwest Fisheries Science Center, 2725 Montlake Boulevard East, Seattle, WA 98112.)

Foerster, R.E. 1968. The Sockeye Sahuon, Bull. Fish. Res. Board Cau. 162, 422 p.

Fraucis, R. C., aud T. H. Sibley. 1991. Climate changes aud fisheries: what are the real issues? Northwest Environ. J. 7:295-307.

Freelaud, H.J. 1990. Sea surface temperatures along the coast of British Columbia regional evidence for a warmiug trend. Cau. J. Fish. Aquat. Sci. 47:346-350.

Healey, M.e. 1980. The ecology of juvenile sahuon in Georgia Strait, British Colunlbia. In W. J. McNeil aud D. C. Hiusworth (eds.), Sahuonid ecosystems of the North Pacific, p. 203-229. Oregon State Univ. Press, Corvallis, OR 97331.

Healey, M.e. 1991. Life history of chiuook sahuon. pp. 313-393 In: C. Groot aud L. Margolis (eds.). Pacific Sahuon Life Histories, UBC Press,

NPAFCBulletinNo.l

Vancouver, B. C., Canada. Heard, W. 1995. An estimate of total 1992 hatchery

releases of juvenile salmon, by country, into waters of the North Pacific Ocean and adjacent seas. (NPAFC Doc.). Auke Bay Laboratory, Alaska Fisheries Science Center, NMFS, NOAA, 11305 Glacier Highway, Juneau, AK. 99801-8626. 6 p.

Hilborn, R., R. Donnelly, M. Pascual, C. Coronado­Hernandez. 1993. The relationship between river flow and survival for Columbia River chinook salmon. Report to Bonneville Power Administration. School of Fisheries, Univ. Washington, Seattle USA.

Isaksson, A. 1988. Sahnon ranching: a world review. Aquaculture 75: 1-33

Johnson, S. 1984. The effects of the 1983 EI Nifio on Oregon's coho sand chinook salmon. Oregon Dep. Fish WildI., Res. Dev. Sec., Portland, Oregon.

Kaeriyama, M. 1996. Population dynamics and stock management of hatchery-reared sahnons in Japan. Bull. Natl. Res. Inst. Aquacult. 2:11-15.

Kaeriyama, M., and S. Urawa. 1992. Future research by the Hokkaido Sahnon Hatchery for the proper maintenance of Japanese salmon stocks. In Y. Ishida, K. Nagasawa, D. Welch, K. Meyers, and A. Shershnev (eds.), Proceedings of the international workshop on future sahnon research in the North Pacific Ocean, p. 57-62. Spec. Pub. Nat. Res. Inst. Far Seas Fish. 20. Shimizu, Japan.

Kemmerich, J. 1945. A review of the artificial propagation and transplantation of the sockeye salmon of the Puget Sound area in the state of Washington conducted by the federal government from 1896 to 1945. Unpublished manuscript, 76 p. plus tables.

Konovalov, S.M. 1980. U.S.S.R: Salmon ranching in the Pacific. In J.E. Thorpe (ed.), Salmon Ranching, p. 63-88, Academic Press, London.

Lawson, P. 1993. Cycles in ocean productivity, trends in habitat quality, and the restoration of sahnon runs in Oregon. Fisheries 18:6-10.

Levings, C.D., C.D. McAllister, and B.D. Chang. 1986. Differential use of the Campbell River estuary, British Columbia, by wild and hatchery­reared juvenile chinook sahnon (Oncorhynchus tshawytscha). Can. J. Fish. Aquat. Sci. 43: 1386-1397.

Lichatowich, J.A. and J.D. Mcintyre. 1987. Uses of hatcheries in the management of Pacific anadromous salmonids. Am. Fish. Soc. Symp. 1: 131-136.

Lichatowich, J. 1993. Ocean carrying capacity recovery issues for threatened and endangered Snake River sahnon, technical report 6 of 11,

Mahnken et al. (1998)

Project 93-013. Available from the Bonneville Power Administration, Portland, OR USA.

Light, J.T. 1989. The magnitude of artificial production of steelhead trout along the Pacific coast of North America. Univ. Washington, Seattle, Fish. Res. Institute pubI. FRI-UW-8913, 11 p.

Mahnken, C.V.W., D.M. Damkaer, and V.G. Wespestad. 1983. Perspectives of North Pacific salmon sea ranching. In Second International Fish Farming Conference, p. 186-216, 22-24 March, 1983, Brighton England. Interbuild Exhibitions Ltd., London.

Mangel, M. and 41 coauthors. 1996. Principles for the conservation of wild living resources. Ecological Applications 6(2):338-362.

Mayama, H. 1985. Technical innovations in chum salmon enhancement with special reference to fry condition and timing of release, 83-86. In C.J. Sinderman (ed.), Proceedings of the eleventh U.S.-Japan meeting on aquaculture, salmon enhancement, Tokyo, Japan, October 19-20, 1982, NOAA Tech. Rep. NMFS 27.

McCabe, G.T., RL. Emmett, W.D. Muir, and T.H. Blahm. 1986. Utilization of the Columbia River estuary by subyearling chinook sahnon. NW Sci. 60(2): 113-124.

McNair, M. 1996. Alaska Fisheries Enhancement Program 1995 Annual Report. Alaska Dep. Fish Game, Reg. Info. Rep. 5J96-08, 43 p.

McNeil, W.J. 1991. Expansion of cultured Pacific salmon into marine ecosystems. Aquaculture 98:173-183.

McNeil, W.J., and J.E. Bailey. 1975. Salmon Ranchers Manual. Northwest Fisheries Center, National Marine Fisheries Service, NOAA, Auke Bay, Alaska, 95 p.

Meffe, G.K. 1992. Techno-arrogance and halfway technologies: salmon hatcheries on the Pacific Coast of North America. Conserv. BioI. 6(3):350-354.

Mullan, J.W. 1986. Determinants of sockeye salmon abundance in the Columbia River, 1880s-1982: a review and synthesis. U.S. Fish Wildt. Servo BioI. Rep. 86(12), 136 p.

National Research Council Committee on Protection and Management of Pacific Northwest Anadromous Sahnonids (NRCC). 1996. Upstream: sahnon and society in the Pacific Northwest. National Academy Press, Washington, D.C. 452 p.

Natural Resource Consultants (NRC). 1995. Artificial propagation of anadromous Pacific sahnonids, 1950 to present. Contract reports to U.S. Dep. Commer., NOAA, NMFS. Includes electronic databases.

NeWsen, W., J.E. Williams, and J.A. Lichatowich.

1991. Pacific salmon at the crossroads; stocks at risk from California, Oregon, Idaho, and Washington. Fisheries 16(2):4-21.

Nickelson, T.E. 1986. Influences of upwelling, ocean temperature, and smolt abundance on marine survival of coho salmon (Oncorhynchus kisutch) in the Oregon Production Area. Can. J. Fish. Aquat. Sci. 43:527-535.

Northcote, T., and D. Atagi. 1994. Pacific salmon abundance trends in the Fraser River watershed compared with other British Columbia systems. In Pacific salmon and their ecosystems Conference, Univ. Wash. Press, Seattle.

Olsen, D., and J. Richards. 1994. Inter-basin comparison study, Columbia river salmon production compared to other West Coast production areas, phase II analysis. Report prepared for the Army Corps of Engineers, Portland, Oregon, by Portland State University and The Pacific Northwest Project. 29 p., plus appendices.

Ohkuma, K., and T. Nomura. 1991. An approach to the efficient enhancement of masu salmon through the release of juveniles into streams. In Marine ranching: Proceedings of the seventeenth U.S.­Japan meeting on aquaculture; Ise, Mie Prefecture, Japan. NOAA Tech. Rep. NMFS 102: 151-159.

Pearcy, W.G. 1992. Ocean ecology of North Pacific salmonids. Univ. Wash. Press. Seattle, Wash. 179 p. Richards, J. , and D. Olsen. 1993. Inter­basin comparison study, Columbia river salmon production compared to other West Coast production areas, phase I analysis. Report prepared for the Army Corps of Engineers, Portland, Oregon, by Portland State University and The Pacific Northwest Project.

Ricker, W.E. 1981. Changes in the average size and average age of Pacific salmon. Can. J. Fish.

53

Aquat. Sci. 38: 1636-1656. Roppel, P. 1982. Alaska's Salmon Hatcheries, 1891-

1959. Alaska Historical Commission Studies in History No. 20. Alaska Historical Commission, Juneau. 299 p.

Schmitten, R., W. Stelle, Jr., and R.P. Jones. 1995. Proposed Recovery Plan for Snake River Salmon, 347 p., plus appendices. (Available from National Marine Fisheries Service, 525 N.E. Oregon, Suite 500, Portland, OR USA 97232-2737.)

Simenstad, C.A., K.L. Fresh, and E.O. Salo. 1982. The role of Puget sound and Washington coastal estuaries in the life history of Pacific salmon: an unappreciated function. In: V. S. Kennedy (ed.) Estuarine Comparisons, p. 343-364. Academic Press, New York.

Wahle, R.J., and R.Z. Smith. 1979. A historical and descriptive account of Pacific Coast anadromous salmonid rearing facilities and a summary of their releases by region, 1960-1976. NOAA Technical Report NMFS SSRF-736.

Welch, D.W., A.I. Chigirinsky, and Y. Ishida. 1995. Upper thermal limits on the oceanic distribution of Pacific salmon (Oncorhynchus spp.) in the spring. Can. J. Fish. Aquat. Sci. 52:489-503.

Zaugg, W.S., E.F. Prentice, and F.W. Waknitz. 1985. Importance of river migration to the development of seawater tolerance in Columbia River anadromous salmonids. Aquaculture 51: 33-47.

Zaugg, W.S., J.E. Bodle, J.E. Manning, and E. Wold. 1986. Smolt transformation and seaward migration in O-age progeny of adult spring chinook salmon (Oncorhynchus tshawytscha) matured early with photoperiod control. Can. J. Fish. Aquat. Sci. 43(4)885-888.

Zorpette, G. 1995. So many salmon, but so little. Sci. American, May 1995, p. 21-22.