Pink salmon (Oncorhynchus gorbuscha) are the most abundant of the five Pacific salmon species in Alaska (Rogers 1986). Wild stocks in Prince William Sound have historically demonstrated negligible odd-even year cycle dominance, although other, longer term trends are evident since 1960 (Fig. 1). Adult returns of less than 10 million fish per year increased to nearly 20 million fish in 1979 and for several years following, but returned to lower production levels in 1986. Since that time, only the returns in 1990, 1991, and 1994 have approached or surpassed 10 million fish. Similar decadal-level production cycling has been described for salmonids in Alaska (Hare and Francis 1995).

A large ocean-ranching program for pink salmon was initiated in 1977. Growing hatchery capacity drove increasingly larger releases of juveniles that peaked in 1991 at 616 million fry. Since 1987, hatchery-released rather than wild pink salmon have been the numerically dominant stock in Prince William Sound. Annual percent marine survivals for hatchery fry to adults have ranged between 1.0 and 9.8, and average 4.4 (Alaska Department of Fish and Game, unpublished). Two of the three poorest production years were recent: 1992 and 1993. There is concern that a hatchery program of this size is seriously impacting the wild production in the region, although direct negative effects have not been demonstrated. An elaborately phased fishery attempts to separate harvests on wild and hatchery stocks as the means to protect the wild escapement each year.

Alaska Department of Fish and Game has censused pre-emergent fry populations since 1962 in index streams in Prince William Sound. These measures generally predict adult returns and have been used as the principal area forecast tools for wild stocks. When the wild return per fry density (a proxy for marine survival) is compared with hatchery marine survivals beginning in 1977, it is apparent that years of good, average, or poor survivals are phased similarly (Fig. 2). This relationship supports a contention that common marine factors influence the survival of both populations.

The salmon literature suggests that pink salmon run strength is established each year during early marine residence, perhaps during the first few weeks in the coastal ocean (Parker 1971, Hartt 1980, Healey 1982, Bax 1983, Hargreaves and LeBrasseur 1985). The mechanism for loss is believed to be predation, and the rate of loss is thought to be modified by the growth rate of the fry during this critical period. The slowest growing fish probably experience the highest rates of mortality since they remain at risk longer in the smallest, presumably weakest, life stages. Temperature and food have been implicated as the major factors influencing growth rates (Walters et al. 1978, Mortensen 1983, Healey 1991). Locally, Willette (1985) and Willette and Cooney (1991) found that production levels of odd and even year southcentral Alaska pink salmon stocks are sensitive to fry-year spring-time ocean temperatures, and that the odd brood lines are also influenced by ocean temperatures during the late maturing and adult stages.

Wild juvenile pink salmon in the northern Gulf of Alaska begin emerging into the nearshore tidally mixed zone in late March. The outmigration from natal habitats is usually completed by early June (Taylor 1988). Cooney et al. (1995) demonstrated close correspondence between the timing of wild juvenile pink salmon ocean entry and the timing and duration of a coastal springtime zooplankton bloom in Prince William Sound. This correspondence suggested that fry benefit by emerging during this period. In fact, marine survival estimates from the hatchery program demonstrate that fry released into the bloom perform better than fry released prior to, or after, the peak of zooplankton biomass. Until recently, this observation seemed to confirm food-limited growth dependence for fry, since juvenile pink salmon are immediate consumers of pelagic food in the deep, nearshore waters of the region (Urquhart 1979). However, more recent studies (Willette in press) demonstrate pink salmon fry growth rates (determined from post-release recaptures of wire-tagged fry) are predicted by springtime temperatures, but not by levels of food. This surprising result means that either zooplankton is rarely growth-rate limiting for fry, or that plankton plays some yet-to-be-determined role in modifying hatchery and, presumably, wild stock production.

In 1993, after a week-long blockade of the Alyeska Pipeline Terminal in Port Valdez by fishermen concerned over failing pink salmon and herring production following the Exxon Valdez oil spill, out-of-court settlement funds were made available to pursue ecosystem-level studies of the early life stages of these two commercial species. In April 1994, the Sound Ecosystem Assessment (SEA) program began the first of five intensive years of field and modeling studies. The SEA program attempts a combined bottom-up and top-down approach to describe factors constraining juvenile pink salmon (and herring) survival. The research focus is on oceanographically modified predation loss as the principal factor regulating pink salmon survival in the juvenile stages each year. Mechanisms influencing fry survival are being used to create a series of physical and biological models envisioned as the major research products of the work, and as carry-forward tools for more informed management of Prince William Sound's pink salmon populations.

In developing its program, SEA began with a simple carbon budget for the region. Given previous estimates of primary productivity, and adopting accepted transfer efficiencies, carbon fixed by phytoplankton was distributed to zooplankton and higher consumers. Enough was understood about the growth of juvenile salmon so their forage demand could be computed on the basis of observed growth rates (Cooney 1993). This surprisingly small demand was then increased by a factor of 10 to estimate the consumption for all 0-age fish in the system, including Pacific herring (Clupea pallasi), northern smoothtongue (Leuroglossus schmidti), and walleye pollock (Theragra chalcogramma). The remaining carbon was distributed to older juvenile fishes and apex consumers, including adult fish, birds, and marine mammals (Fig. 3).

The carbon budget suggests that zooplankton should support large populations of consumers, and that the region may also serve as a juvenile fish nursery. The model further suggests that when zooplankton stocks are low—seasonally or annually—intermediate and large-size consumers can augment their diets by eating more smaller fishes. Since the subarctic surface zooplankton community is greatly diminished during the fall and winter months, the pelagic system is probably more planktivorous during the spring and summer months, and piscivorous during much of the remainder of the year. The present research program focuses on the role that zooplankton and other factors play in modifying losses of juvenile salmon to predators during the critical first few weeks of early marine residence. Field studies take advantage of large fry releases from hatcheries to observe and measure local predation losses on timescales of days to weeks during a sampling period from April through mid-June.

After 2 years of study, the SEA program has begun to assemble an understanding of processes regulating the survival of wild and hatchery released juvenile pink salmon. Fry predator studies have demonstrated that birds (principally kittiwakes) and walleye pollock play a major role in fry losses each year. Personnel of local hatcheries have been aware of bird and pollock predation, but they have never attempted a quantitative assessment of the annual impact. Midwater trawling and acoustic surveys conducted near hatcheries in western Prince William Sound during the spring and summer of 1994 and 1995 established that adult pollock from a successful 1988 year class were the most dominant members of the nearsurface large-fish community from late April through early July. In April and May, a significant percentage of the diet of these fish was zooplankton; calanoid copepods were dominant and represented mostly by Neocalanus spp. During both 1994 and 1995, large numbers of calanoids occurred concurrently in the water column and stomachs of the walleye pollock. In both years, the consumption of 0-age fishes, including juvenile salmon, declined or remained at low levels in pollock diets during the copepod bloom (Fig. 4). Since wild and hatchery stocks of juvenile salmon migrate across deep water each year as they move southward in the region, offshore losses to adult pollock are probably modified by amounts of zooplankton present at this time.

In addition to salmon fry being vulnerable to adult walleye pollock over the deeper passages, more recent results suggest that in late May and early June of some years, 1- and 2-year-old pollock move into nearshore tidally mixed zones to feed. This is a time when most fry have grown to 50-70 mm in size and are leaving the shallow waters. Stomach analyses indicate that some juvenile pollock feed heavily on salmon fry. Depending on the time phasing and numbers of each year, these late-arriving predators could account for substantial losses of juvenile pink salmon, particularly during years when reduced fry growth may extend their nearshore nursery period. The significance of fry size in relation to predation was examined in 1994 by experimentally releasing some fry that had been grown to over 1-g live weight. These fish, approximately four times larger than fry normally released from hatcheries, exhibited marine survivals an order of magnitude higher than their smaller surviving siblings (Table 1).

Other predators, including adult herring, squid, greenling (Hexagrammos spp.), Dolly Varden trout (Salvelinus malma), and adult salmon, also prey on juvenile salmon. With the possible exception of squid, these other predators seem to play a lesser predation role than walleye pollock. Unfortunately, there are problems associated with quantitatively sampling most of these other species. Squid are probably not taken adequately in trawls or fixed gear, and the nearshore regime cannot be fished with trawls because of the steep and rocky local topography. Seines, gill nets, and traps are providing some qualitative data, and side-scan sonar may help in the future.

Surprisingly, the monthly tidal cycle is also implicated in the survival of juvenile salmon. Survival is better under spring-tide rather than neap-tide conditions (Fig. 5). Observations this past year also demonstrate that catches of adult pollock were statistically correlated with tidal state. At the height of a mid-May spring-tide series, adults left the upper layers only to reappear later as the spring cycle relaxed (Fig. 6). We surmise that increased currents and turbulence associated with larger tidal lenses cause fish populations to reorient their distributions, and that feeding may be curtailed during these times.

In summary, a combined oceanographic and fisheries study in Prince William Sound is beginning to describe factors that interact to modify predation losses in juvenile salmon populations each year. Generally, pink salmon seem to have evolved a life history strategy that brings the juveniles into marine waters during an intense period of plankton growth each spring. This timing—on average—provides forage and seasonally warming temperatures that probably stimulate critical early season growth rates. The presence of immense upper-layer calanoid populations also provides food for most other consumers during this same time. Many fish and birds that eat fry find energy in abundant plankton populations as well. For any given year, the time-space phasing of fry, plankton, and consumer fish and bird populations, together with ocean temperatures, probably establish the conditions that modify predation on juvenile pink salmon. Warm years with high zooplankton stocks should enhance survival of rapidly growing fry that escape predation because most of the energy flowing to larger fishes, birds, and mammals comes from plankton. In contrast, cold years with reduced plankton biomass should be disastrous to fry whose growth rates would be reduced and who would be preyed upon more heavily by larger consumers unable to obtain sufficient energy from plankton. These extreme examples occur in the historical salmon record. However, there are enough years when fry growth rate and plankton sheltering alone fail to predict production to negate the applications of just this information.

Because of inherent system complexities, it is unlikely that our overall results will ever lend themselves to highly accurate survival estimates of pink salmon based on fry year growth conditions and predation loss factors. However, we do expect that simulations of major parts of the system will provide reasonably consistent projections of good, average, or poor marine survivals one year in advance of adult returns. This consistency is presently lacking in forecasts of pink salmon in Prince William Sound, to the detriment of local harvest-management and enhancement programs.

To accomplish this goal, the SEA program is proposing a nominally funded long-term monitoring program to reinitialize the model(s) for selected bottom-up forcing (temperature, winds, fresh water, other), for validation purposes, and to index, rather than predict, predator populations each year. The SEA program is expected to complete by 1997 much of the modeling and to have a working, constantly updated, forecasting model running by 1999.

Bax, N. J. 1983. Early marine mortality of marked juvenile chum salmon (Oncorhynchus keta) released into Hood Canal, Puget Sound, Washington, in 1980. Can. J. Fish. Aquat. Sci. 40:426-435.

Cooney, R. T. 1993. A theoretical evaluation of the carrying capacity of Prince William Sound, Alaska, for juvenile Pacific salmon. Fish. Res. 18:77-87.

Cooney, R. T., T. M. Willette, S. Sharr, D. Sharp, and J. Olsen. 1995. The effect of climate on North Pacific pink salmon (Oncorhynchus gorbuscha) production: Examining some details of a natural experiment. In R. J. Beamish (editor), Climate change and northern fish populations. Can. Spec. Publ. Fish. Aquat. Sci. 121:475-482.

Hare, S. R., and R. C. Francis. 1995. Climate change and salmon production in the Northeast Pacific Ocean. In R. J. Beamish (editor), Climate change and northern fish populations. Can. Spec. Publ. Fish. Aquat. Sci. 121:357-372.

Hargreaves, N. B., and R. J. LeBrassuer. 1985. Species selective predation on juvenile pink salmon (Oncorhynchus gorbuscha) and chum salmon (O. keta) by coho salmon (O. kisutch). Can. J. Fish. Aquat. Sci. 42:659-668.

Hartt, A. C. 1980. Juvenile salmon in the oceanic ecosystem—the critical first summer. In W. J. McNeil and D. C. Himsworth (editors), Salmonid ecosystems of the North Pacific. Oregon State University Press, Corvallis, p. 25-57.

Healey, M. C. 1982. Timing and relative intensity of size-selective mortality of juvenile chum salmon (Oncorhynchus keta) during early sea life. Can. J. Fish. Aquat. Sci. 39:952-957.

Healey, M. C. 1991. Diets and feeding rates of juvenile pink, chum and sockeye salmon in Hecate Strait, British Columbia. Trans. Am. Fish. Soc. 120:303-318.

Mortensen, D. G. 1983. Laboratory studies on factors influencing the first feeding of newly emerged pink salmon (Oncorhynchus gorbuscha) fry. M.S. Thesis, 108 p. Univ. Alaska, Juneau.

Parker, R. R. 1971. Size selective predation among juvenile salmonid fishes in a British Columbia inlet. J. Fish. Res. Board Can. 28:1503-1510.

Rogers, D. E. 1986. Pacific salmon. In D. W. Hood and S. T. Zimmerman (editors), The Gulf of Alaska: Physical environment and biological resources. U.S. Dep. Comm., NOAA, Environ. Res. Lab., p. 461-478. (Available from Arctic Environ. Assess. Cent., 222 W. 8th Ave., No. 56, Anchorage, AK 99513.)

Taylor, S. G. 1988. Inter- and intra-annual survival of pink salmon (Oncorhynchus gorbuscha) returning to Auke Creek, Alaska in 1986 and 1987. APPRISE Annual Report 1987. School of Fisheries and Ocean Sciences, Univ. Alaska, Fairbanks, Tech. Rep. 1:547-571.

Urquhart, D. L. 1979. The feeding, movement, and growth of pink salmon (Oncorhynchus gorbuscha) fry released from a hatchery in Prince William Sound, Alaska. M.S. Thesis, Univ. Alaska, Fairbanks, 111 p.

Walters, C. J., J. Holborn, R. M. Peterman, and M. J. Staley. 1978. Model for examining early ocean limitation on Pacific salmon production. J. Fish. Res. Board Can. 35:1303-1315.

Willette, T. M. 1985. The effects of ocean temperatures on the survival of odd- and even-year pink salmon (Oncorhynchus gorbuscha) populations originating from Prince William Sound, Alaska. M.S. Thesis, Univ. Alaska, Fairbanks, 115 p.

Willette, T. M. In press. Impacts of the Exxon Valdez oil spill on the migration, growth and survival of juvenile pink salmon in Prince William Sound, Alaska. Proceedings of EVOS Symposium, Feb. 1993, Anchorage, Alaska.

Willette, T. M., and R. T. Cooney. 1991. An empirical orthogonal functions analysis of sea surface temperature anomalies in the North Pacific Ocean and cross correlations with pink salmon (Oncorhynchus gorbuscha) returns to southern Alaska. In B. White and I. Guthrie (editors), Proceedings of 15th Pink and Chum Salmon Workshop. Pacific Salmon Commission, p. 111-121. (Available from Pacific Salmon Commission, 600-1155 Robson St., Vancouver, BC, Canada V6E 1B9.)

Table 1. Marine survivals for hatchery pink salmon by release size and date for two hatcheries in Prince William Sound, 1994.

A.F.K. = Armin F. Koernig Hatchery.

W.H.N. = Wallace H. Noerenberg Hatchery.

Table of Contents