Auke Bay Lab Quarterly Research Report for Oct-Dec '99

The final sablefish assessment was prepared by members of the Auke Bay Laboratory (ABL) and Resource Ecology and Fisheries Management (REFM) Division. The assessment shows that sablefish abundance increased during the mid-1960s due to strong year classes from the late 1950s and 1960s. Abundance subsequently dropped during the 1970s due to heavy fishing; catches peaked at 56,988 metric tons (t) in 1972. The population recovered due to exceptional year classes from the late 1970s; spawning abundance peaked again in 1987. The population then decreased as these exceptional year classes died off.

The longline survey abundance index increased 10% in numbers and 5% in weight, and the commercial fishery abundance index increased 11% in weight from 1998 to 1999. These increases follow decreases from 1997 to 1998, so that relative abundance in 1999 is similar to 1997. Exploitable and spawning biomass are projected to increase 3% and 1%, respectively, from 1999 to 2000. Alaska sablefish abundance now appears low and stable. This is a change from previous assessments where abundance appeared low and slowly decreasing. Further years’ data are needed to confirm that abundance has stabilized.

A simple Bayesian analysis was completed by examining the effect of uncertainty in natural mortality and survey catchability on parameter estimation. A decision analysis was completed using the posterior probability from the Bayesian analysis to determine what catch levels likely will decrease abundance. The decision analysis indicates that a yield of about 17,000 t most likely will keep spawning biomass the same and has only a 20% probability of reducing the 2004 spawning biomass to less than 90% of the 2000 spawning biomass. The maximum permissible yield from an adjusted F40% strategy is 17,300 t, which was the 2000 Acceptable Biological Catch (ABC) accepted by the North Pacific Fishery Management Council (NPFMC) for the combined stock, a 9% increase from the 1999 ABC of 15,900 t.

Stock assessment for major groundfish stocks in the Bering Sea, Aleutian Islands, and Gulf of Alaska are presented in this issue in the REFM Division’s report in this section. Stock assessments of slope and pelagic shelf rockfish in the Gulf of Alaska follow.

By Michael Sigler.

Stock Assessment of Slope and Pelagic Shelf Rockfish in Gulf of Alaska

Updated stock assessments of slope rockfish and pelagic shelf rockfish in the Gulf of Alaska were completed by ABL and REFM scientists. Results of the 1999 NMFS triennial trawl survey were included in the updated stock assessments. The completed assessment of Pacific ocean perch (POP), a member of the slope rockfish assemblage, used an age-structured model which showed that the stock is increasing with an estimated exploitable biomass of 200,310 t. The assessment of most other species of slope rockfish and pelagic shelf rockfish in the Gulf of Alaska rely on biomass estimates provided by trawl surveys, and abundance trends based on these surveys are highly uncertain. The most recently completed assessments indicate the following stock levels (exploitable biomass): shortraker and rougheye rockfish, 70,890 t; northern rockfish, 85,360 t; other slope rockfish, 102,510 t; pelagic shelf rockfish, 66,440 t. Recommended ABC levels were Pacific ocean perch,13,020 t; shortraker and rougheye rockfish, 1,730 t; northern rockfish, 5,120 t ; other slope rockfish, 4,900 t; and pelagic shelf rockfish 5,980 t. These ABC values were all accepted by the NPFMC.

A new age-structured assessment model for northern rockfish was presented at the November NPFMC Gulf of Alaska groundfish plan team meeting. The model incorporates data from several disparate fisheries and surveys and provides a quantitative framework for evaluating the influence of different data sources on the estimated population. The model also provides a suggested rationale for determining an appropriate level of confidence in apparently inconsistent data components. Since an initial review in September, the model was updated to include 1999 survey biomass and length compositions, 1999 catch, and 1996 survey age compositions. The model was also revised to include several modifications suggested by the GOA plan team and by the NPFMC Scientific and Statistical Committee. A report describing the model was included as an appendix in the 1999 stock assessment fishery evaluation (SAFE) document. The model is expected to be used for the determination of northern rockfish stock status for the 2001 fishery.

By Dean Courtney and Jon Heifetz.

Eight Tagged Sablefish Recovered in 1999

ABL scientists began tagging sablefish in 1998 with electronic tags that record depth and temperature. (See feature article “New Sablefish Research at the Auke Bay Laboratory” in the July-September 1999 issue of the AFSC Quarterly Report.) Recovery of the tags will yield new insight into sablefish daily, seasonal, and age-related depth movements and the marine environmental conditions in which they live. Knowledge of these movements will lead towards better recommendations of sustainable harvests by understanding what part of the population is susceptible to the fishery and how this susceptibility changes during the life of the fish.

During the 1998 NMFS longline survey of the Aleutian Islands region and Gulf of Alaska (GOA), electronic tags measuring 3/4 inches diameter by 2 1/4 inches long were surgically implanted in the abdominal cavity of 195 sablefish. The fish also were externally marked with a flourescent pink and green tag. More tagging is planned for the 2000 longline survey.

Nine electronic tags have been recovered so far, one during 1998 and eight during 1999. One tagged fish was at large 2 months, and the remainder about 1 year. Most fish were recovered near their release location, except for one fish that traveled from near Dutch Harbor to near Seward. Average weekly depth for each recovered fish ranged from 200 to 500 m during summer and 400 to 700 m during winter. Average weekly temperature ranged from 3.5o to 6.5oC, but no seasonal pattern was apparent. Individual fish traveled a wide depth range, sometimes during 1-2 days. For example, one fish rose from about 1,100 m to 670 m in 1-1/2 days and to 220 m in 9 days during October 1998.

By Michael Sigler.

Alaska Longline Survey Completed

On 5 September 1999, AFSC scientists completed the twenty-first annual longline survey of the upper continental slope of the Gulf of Alaska and a portion of the slope of the eastern Bering Sea. One hundred-fifty-two longline hauls (sets) were completed. Sablefish was the most frequently caught species, followed by giant grenadiers, Pacific cod, arrowtooth flounder, and Pacific halibut. A total of 88,949 sablefish, with an estimated total round weight of 298,146 kg (657,412 lb), was taken during the survey. The highest total sablefish catch was observed at station 105 in southern Southeast Alaska. Station 98 in northern southeast Alaska had the largest average length for sablefish.

A total of 4,633 sablefish, 603 shortspine thornyhead, and 188 Greenland turbot were tagged and released during the survey. Length-weight data and otoliths were collected from 2,451 sablefish. Thirty-six surface gillnet sets were completed to assess the abundance of juvenile sablefish. A very low number of sablefish (28 young-of-the year-and 12 age-1) were caught in the gillnets during the 1999 survey.

Killer whales preying on sablefish and Greenland turbot caught on the gear were observed at seven eastern Bering Sea stations.

By Thomas Rutecki.

Effects of Trawling on Soft-bottomed Marine Habitat

An 11-day research cruise to study the effects of bottom trawling on the seafloor near Kodiak Island was completed on 24 August. The cruise used the manned submersible Delta and the Alaska Department of Fish and Game (ADF&G) vessel Medeia and was the final cruise of a 2-year study to make observations of the seafloor in areas open to bottom trawling and adjacent areas that have been closed to bottom trawling since 1986. Processing of all sediment samples for infaunal composition, grain size composition, and organic carbon content were completed during the quarter. Analysis of video footage taken from the submersible, a labor intensive task, is continuing, and fish and invertebrate counts for all sites should be completed by October 2000, and project analyses by the end of 2000. An expanded summary of this project was provided in the July - September 1999 issue of the AFSC Quarterly Report. Summaries of this and other AFSC studies on the effects of fishing may be viewed at the following Internet address: http://www.afsc.noaa.gov/abl/MarFish/pdfs/heifhabitatdec99.pdf.

By Robert Stone.

Alaska Chum Salmon Issues

A special meeting to discuss current chum salmon (Oncorhynchus keta) production trends in Alaska was hosted by Douglas Island Pink and Chum Salmon, Inc. (DIPAC) at the Gastineau Hatchery in Juneau, Alaska, on 5-6 January. Attendees included representatives of the ADF&G, University of Alaska School of Fisheries and Ocean Sciences, Southern Southeast Regional Aquaculture Association of Ketchikan, Northern Southeast Regional Aquaculture Association of Sitka, Prince William Sound Aquaculture Association of Cordova, Trident Seafoods, Icicle Seafoods, Norquest Seafoods, Taku Fisheries, Ward’s Cove Packing, and scientists from the ABL.

An objective of the meeting was to review recent returns of chum salmon throughout Alaska, with a focus on hatchery returns, especially in Southeast Alaska. The aquaculture associations and DIPAC gave historical overviews of recent returns along with analyses of their methodologies for forecasting expected returns to their facilities in 2000. Recent commercial harvests of chum salmon in Alaska have approached or exceeded historical high levels. The abundance of chum salmon, in turn, has stretched the capabilities of traditional processing and marketing methods for this species. The statewide commercial harvest of chum salmon in 1998 and 1999 exceeded 19.0 and 20.0 million fish, respectively, with about 75% of the catch in the southeast region. A majority of these fish were hatchery-produced.

Industry representatives indicated they were pleased with the new quantities of chum salmon but were interested in hearing from aquaculture associations, scientists, and managers about the prospects for continued large returns. They emphasized a critical need for long-range planning to develop adequate outlets for various qualities of chum salmon product, including bright- and dark-fleshed fish as well as different grades of egg product. They pointed out that most Alaska chum salmon eggs are processed as salmon caviar for Japanese markets and that the eggs are the source of Alaska chum salmon’s commercial value. Developing products, markets, and outlets for Alaska chum salmon, including dark and maturing fish, takes concerted effort, patience, and time in finding and developing economically viable solutions to current production levels.

The meeting also focused on recent age structure patterns in chum salmon to determine if large returns of age-3 fish are a precursor to large returns of age-4 fish the following year. Results of the review indicated that this relationship was generally true, and aquaculture associations are predicting returns to Southeast Alaska hatcheries in 2000 of 14.0-16.0 million chum salmon, similar to the last 2 years.

Jack Helle, from the ABL, reviewed age structure, survival, and size-at-age patterns in wild stocks of chum salmon based on a 25-year database of samples throughout different parts of their North American range. Until recent years, there has been a general decline in size-at-age of chum salmon as abundance of chum salmon has increased, which has been interpreted as evidence of density-dependent interactions and carrying capacity limitations in ocean feeding areas. However, size-at-age has increased dramatically for Southeast Alaska and Pacific Northwest chum salmon in the past few years, concurrent with historical record catches. Helle also reviewed ABL’s Ocean Carrying Capacity (OCC) Program current research on sampling juvenile salmon during their first year at sea. He noted that much new information on migration and distribution of juvenile salmon is now possible due to the otolith marking technology used by several hatcheries in Alaska and elsewhere.

Discussions moved to variations in ocean survival and size-at-age caused by regime shifts in climate patterns and the large differences in survival patterns and run strength in different regions throughout the Pacific Rim. Examples included a recent decline in chum salmon returns in Hokkaido and Honshu compared to the many Alaska returns that remain strong. A notable exception to the strong upward trend for chum salmon in Alaska is the extremely poor return in the Arctic-Yukon-Kuskokwim (AYK) region. Reasons for the poor returns in AYK are unknown and could relate to many factors either in fresh water or at sea or both. Southeast Alaska chum salmon have been identified in salmon bycatch in the Bering Sea, which has prompted some speculation that the high abundance of chum salmon from Southeast Alaska and other regions could be negatively influencing the survival of western Alaska populations. These by-catch recoveries have all been of immature (ocean age 1-2) fish. However, juvenile salmon (ocean age 0) captured in July and September 1999 during OCC research cruises in Bristol Bay and the southern Bering Sea were predominately sockeye salmon; no juvenile chum salmon were captured. Previous OCC cruises at this time of year in the Gulf of Alaska have found juvenile chum salmon from southeast Alaska distributed in a coastal band in the Gulf of Alaska, from Cape Spencer to the Alaska Peninsula. These observations suggest that there is little mixing of the western AYK juveniles with the southern stocks during their first ocean year, the presumed critical period for determining marine survival. Interactions of older age groups are generally thought to influence growth and size and age at maturity, but not survival. Understanding where and how AYK chum salmon spend their first year at sea could provide important clues to some of the causes for poor returns to that region. Although considerable research has focused on Asian juvenile salmon that enter the Bering Sea, unfortunately very little is known about young salmon entering the Bering Sea from the AYK Region.

By Bill Heard and Alex Wertheimer.

Origins of Sockeye and Chum Salmon Seized From the Vessel Ying Fa

Samples of chum and sockeye (O. nerka) salmon seized from the stateless fishing vessel Ying Fa were analyzed to determine their region of origin using genetic stock identification (GSI), otolith marks, parasite analysis, and scale data. Based on GSI, the chum salmon samples originating in Russia were 86%; Japan, 2%; western Alaska, 2%; Alaska Peninsula and Kodiak, 8%; and British Columbia, 2%. Origins of the sockeye salmon sample were not as clear because there was some disagreement between the parasite data and the GSI and scale data. Results of parasite analysis suggested the sample was nearly all of Alaskan origin, with at least 15% coming from Bristol Bay. The GSI analysis indicated that 30% of the sockeye salmon originated in Russia and 70% in North America. The scale analysis showed that 97% of the sockeye salmon sample were ocean age 3, whereas the return to Bristol Bay in 1999 was approximately 70% ocean age 2 fish. This report was submitted to the North Pacific Anadromous Fish Commission (NPAFC) and is available as a pdf file on the AFSC web site at http://www.afsc.noaa.gov/abl/StockID/pdfs/yingfa.pdf

By Richard Wilmot.

Comprehensive Allozyme Database Discriminates Chinook Salmon Around Pacific Rim

The Chinook Salmon Genetics Working Group, which is comprised of four west coast genetics laboratories, met in Anchorage on 4-6 October 1999 to complete a coast-wide genetic database of 265 chinook salmon populations. Allozyme allele frequencies from numerous studies have been standardized and combined into a single comprehensive database, which is managed by the Northwest Fisheries Science Center (NWFSC). A report was submitted to the NPAFC which documents the collaborative database constructed at that meeting and the results of simulations conducted to identify genetic groups that can be accurately and precisely identified in mixtures. Previous versions of the allozyme database included nearly 200 populations ranging from the Sacramento River in California to the Stikine River in British Columbia. However, adequate coverage for Alaska and Russia had been lacking, so no Alaskan, Russian, or high-seas applications were possible. In recent years, the ADF&G and the ABL have initiated programs to update, enlarge, and standardize allozyme data from northern and western Alaska populations to develop a species-wide database. Additional populations from British Columbia as well as other Pacific Northwest populations have also become available.

By Charles Guthrie.

Yukon River Radio-Tagging Project

The ABL’s radio-tagging study of Yukon River fall chum salmon was completed in fall 1999 with over 1,000 fish radio-tagged and tracked to their final destination. Fall aerial surveys were conducted to locate fish in areas associated with the Yukon River main stem.

Work continues on 1) planning for telemetry studies on Yukon River chinook salmon and 2) completion of the Proceedings of the 15th International Symposium on Biotelemetry hosted by NMFS in Juneau during May 1999.

By John Eiler.

Lipid Class and Fatty Acid Analysis of Forage Fish Diets

Biologists at the ABL have been developing lipid class and fatty acid analysis as a tool for examining trophic relationships in forage fish species. Surplus lipids acquired in a fish’s diet are stored as triacylglycerides (TAG). Consequently, the fatty acid (FA) composition of a fish’s TAG should reflect that of its diet. In addition, the amount of TAG in a fish relative to the total lipid should provide a measure of its energetic reserves because the FAs stored in TAG are ultimately used to fuel metabolism. We have been evaluating the use of these measures for detecting dietary differences in forage fish collected in Prince William Sound (PWS) and evaluating the quality of the diets.

Our initial examination was to determine if there was a spatial component to variation in FA composition. We collected juvenile herring and sand lance samples from several locations in PWS. All collections were made within a 10-day period to reduce any temporal variation. Initial analysis revealed statistically significant differences in the TAG content of herring and sand lance collected at different locations. The highest TAG levels were observed when both species co-occurred.

We compared the FA compositions of the TAG using a supervised classification tool that devises a set of rules for classifying samples into groups defined by the user. Once the rules have been devised, they are validated with samples whose types are known, but were not used to devise the rules. The classification tool provided a model for classifying species by their FA compositions that was accurate more than 95% of the time, and one that correctly classified the sampling location more than 85% of the time. However, the spatial model failed to make any distinction between species at a given location. This suggests that there is an important spatial component to the variation in FA composition, and it is likely that this results from similarities in the diets of the fish at a given location regardless of species.

Work under way will compare the FA compositions of the herring and sand lance to zooplankton sampled at the same time, to determine if spatial differences in FA composition are influenced by prey type. In a related study we plan to examine the lipid class and FA composition of sand lance sampled from a single site every 2 weeks between May and September 1998.

By Ron Heintz.

Evidence and Consequences of Persistent Oil in Pink Salmon Streams

Water of some Prince William Sound (PWS) salmon spawning steams remains contaminated with Exxon Valdez oil a decade after the spill (fall 1999). Total aqueous polynuclear aromatic hydrocarbon (PAH) concentrations were highest (300-400 ng/g) in the lower intertidal area in two of six previously oiled streams, supporting a hypothesis that oil remaining in adjacent beaches can contaminate stream water.

Persistent oil contamination may have inhibited recovery of wild pink salmon stocks through the mid-1990s. Reduced viability of pink salmon embryos in oiled PWS streams through 1993 was reported by Bue et al. (1998). This observation, however, has been contested by others (Brannon and Maki 1996). Thus, a decade after the 1989 Exxon Valdez oil spill, the impact of the spill on wild pink salmon populations remains problematic and controversial. As a whole the pink salmon population in PWS has been healthy for some time, but there were reports of persistent oil contamination in stream banks through 1995.

Central to the embryo viability controversy is whether any salmon streams in PWS were ever oiled.   Flowing water did not allow oil to float upstream in 1989, thus oil was stranded on stream banks, deltas, and adjacent intertidal beaches, but not on gravel in stream beds. However, a mixed function oxidase enzyme, cytochrome P4501A, was induced in alevins from oiled streams but not in those from reference streams. Induction of this biomarker enzyme signals exposure to PAH or halogenated hydrocarbons, suggesting that pre-emergent fish were exposed to oil.

Accumulating laboratory evidence demonstrates that exposure of incubating pink salmon eggs to PAH dissolved from oil reduces embryo survival, growth rates, predator avoidance, and marine survival. Exposure of developing pink salmon eggs to oil was first verified as a plausible explanation for reduced viability in 1992 by incubating pink salmon eggs in gravel coated with known amounts of oil and has been verified by repeated laboratory experiments in Southeast Alaska.

Contamination of PWS stream water by PAH may explain elevated pink salmon embryo mortality in oiled streams from 1989 through 1993, but the mechanism for oil transfer from surrounding gravel to stream water was unknown. Coupled with the observation that PAH dissolve from oiled rock into water at concentrations sufficient to cause toxicity, a hypothetical mechanism to explain toxic levels of PAH in stream water was devised.   This hypothesis suggests that PAH from oil on or in stream banks or adjacent intertidal beaches will dissolve in interstitial water, and that a portion of contaminated interstitial water, driven by the dynamics of tidal fluctuations and hydraulic gradients, will enter surface or subsurface stream water. Lipophylic salmon eggs buried in stream gravel could then be contaminated by PAH even though gravel in stream beds remains uncontaminated.

Current research is directed at verification of the PAH-transfer hypothesis, and interpretation of P4501A induction. The recent detection of PAH in PWS stream water with lipophylic samplers is evidence that oil in stream banks can contaminate stream water. Further hydrographic research is planned to further test the PAH-transfer hypothesis. Graded oil-dose laboratory experiments with pink salmon eggs are underway to link induction of the P4501A biomarker to marine survival.

By Mark Carls.

NPAFC Annual Meeting and Symposium

The Auke Bay Laboratory cohosted the Seventh Annual Meeting of the North Pacific Anadromous Fish Commission (NPAFC) held on 24-29 October 1999 in Juneau, Alaska. The NPAFC was established by the Convention for the Conservation of Anadromous Stocks in the North Pacific Ocean (the Convention) which entered into force on 16 February 1993. The Convention prohibits directed fishing for salmonids on the high seas of the North Pacific Ocean and includes provisions to minimize the number of salmonids taken in other fisheries. The NPAFC promotes the conservation of salmonids in the North Pacific and its adjacent seas and serves as a venue for cooperation in and coordination of enforcement activities and scientific research. The commission is made up of representatives of Canada, Japan, Russia, and the United States.

Jack Helle of the ABL chaired the U.S. section of the NPAFC Committee on Scientific Research and Statistics which reviewed and discussed scientific research on a broad range of issues concerning Pacific salmonid stocks, including the relationship between changes in abundance and in ocean and atmospheric conditions and other biological and ecological dynamics of salmonid production. Following the annual meeting, salmon scientists met on 1-2 November in Juneau for a 2-day symposium entitled “Recent Changes in Ocean Production of Pacific Salmon.” Keynote speakers were Dr. Elbert (Joe) Friday from the National Academy of Sciences and Dr. Bruce Finney from the University of Alaska. Eight presentations submitted by ABL staff were selected for publication in the symposium proceedings.   Abstracts of the papers are given below.

On chum salmon (Oncorhynchus keta) scales, we compare the reference line extending from the focus to the edge along an axis drawn 75E from the transition zone (INPFC (International North Pacific Fisheries Commission) scale method) to the anterior-posterior line (traditional scale method) by stock, brood year, and age. Fish lengths calculated from the two scale axes differed significantly (p = 0.004) for all stocks, brood years, and ages except age-5 fish from Fish Creek, Alaska, brood year 1981 (p = 0.27). Mean calculated fish length at ocean age was greater for the traditional scale method by 0%-4% (3-6 mm), 0%-4% (1-9 mm) at ocean age 2, 0%-2% (2-10 mm) at ocean age 3, and 0%-2% (0-10mm) at ocean age 4.

Scale proportions for the two axes differed significantly (p = 0.003) for all stocks, brood years, and ages except age-5 fish from Fish Creek, brood year 1981 (p = 0.28). Scale proportions differed significantly between the two reference lines; however, the absolute differences in proportions were small. Scale measurements were consistently greater for the INPFC scale method in all years of marine growth: by 1%-9% in the first year, 6%-13% in the second year, 0%-14% in the third year, 0%-14% in the fourth year, and 7%-24% in the last year. Scale measurements were consistently different by brood year and age, but varied by stock.

We examined variation in time of annulus formation on scales (n = 552) of chum salmon collected in the North Pacific Ocean in April 1999 and May 1998 and 1999. May samples were collected from approximately the same location and same time, but during 2 years of very different ocean conditions: 1998, a strong El Niño year and 1999, a strong La Niña year. Annuli occurred during April and May and were formed earlier in 1999. Number of circuli beyond the last annulus revealed an inverse relationship to age-1, -2, -3, and -4 fish, but not for age-5 fish, indicating that annulus formation takes place earlier in younger fish. This was verified by our findings. In most younger fish (age-1, -2, and -3) the annulus was completed in April, and in older fish (age-4 and -5) the annulus in many fish was not completed by the end of May.

Immature sockeye salmon (Oncorhynchus nerka) were collected in the coastal waters of the eastern Bering Sea (Cape Cheerful; Unalaska Island) during April/May and in the coastal waters of the Gulf of Alaska (Cape Prominence; Unalaska Island) during August 1998. Genetic stock identification techniques (protein electrophoresis) indicated that Bristol Bay stocks made up the largest percentage in the samples. The majority of samples of immature sockeye salmon collected during April/May (n = 430) and August (n = 300) were age 1.1. Stomachs of immature sockeye off Cape Cheerful contained mostly fish, and off Cape Prominence mostly pteropods. The substantial number of immature sockeye salmon captured at Cape Cheerful during May 1998 was unexpected based on current migration models of western Alaska sockeye salmon. The large percentage of immature sockeye salmon (98%) captured at Cape Prominence during August 1998 was also unexpected since immature chum salmon (O. keta) constituted the largest percentage of the immature salmon catch during August 1996 at the same location (77%) (Carlson et al., 1996) and 1997 (66%) (Carlson et al., 1997). These unexpected events may be due to changes in distribution resulting from the strong El Niño event during 1997-98.

Spatial variations in early marine growth and condition were examined for hatchery-raised juvenile pink (Oncorhynchus gorbuscha) and chum (O. keta) salmon collected in the coastal waters of the Gulf of Alaska (GOA) during 1996. Hatchery salmon released in spring 1996 were recovered between 25 July and 9 August 1996. Most pink salmon from Prince William Sound hatcheries were distributed southwest along the continental shelf from Cape Puget to Mitrofania Island along the Alaska Peninsula. Chum salmon from southeastern Alaska hatcheries were distributed northwest along the continental shelf from Cape Spencer to Cape Hinchinbrook. In general, mean lengths and weights of juvenile pink salmon from Prince William Sound hatcheries and juvenile chum salmon from southeastern Alaska hatcheries increased as fish migrated westward along the coast; the smallest individuals were found near the exit corridors where these salmon first entered the GOA. Condition factor was lowest for salmon caught nearest these exit corridors but increased as fish migrated westward along the coast.

We report the results of broad-scale, shipboard surveys of ocean distribution and migration patterns of Pacific salmon (Oncorhynchus spp.) in coastal and offshore waters of the North Pacific Ocean in spring and summer 1996-99. A large, midwater rope trawl was used to catch salmon and enhanced our ability to sample broad oceanic areas in relatively short periods of time, even under marginal weather and sea conditions. Recoveries of hatchery salmon with thermal otolith marks provided new stock-specific data on the distribution of central and southeastern Alaskan pink (O. gorbuscha) and chum (O. keta) salmon in the Gulf of Alaska. In general, the survey results corroborate the findings of previous studies of ocean distribution and migration patterns of salmon. New information indicates that: 1) off-shore distribution of juvenile (ocean age .0) pink and chum salmon varies by geographic region, which may reflect differences in the width of the continental shelf; 2) a few juvenile southeastern Alaska hatchery chum salmon were caught south of major exit corridors, counter to the predominant northward migration pattern; 3) the northern Shelikof Strait may be an important summer migration corridor for juvenile pink salmon, 4) the ocean range of central Alaska pink salmon extends further to the southwest (to 42°N, 165°W) than shown by high-seas tag experiments; and 5) some maturing chum salmon caught in the coastal waters off Prince William Sound in May are from an early southeastern Alaska hatchery run (peak harvest in mid-July). We conclude that sufficient numbers of thermally-marked hatchery salmon can be recovered during coastal and offshore salmon surveys to provide significant new stock-specific information on ocean distribution and migration patterns of salmon.

Growth measurements were taken from 9,414 legible scales of Kvichak, Bristol Bay, sockeye salmon (Oncorhynchus nerka), yielding a long time series (1914-97) of ocean growth data. Growth rates in the first, second, and third ocean years all declined prior to the late 1950s and early 1960s after which they began to steadily increase until 1970 when the three growth patterns diverged: first year growth continued to increase, but at a lower rate; second year growth showed no further increase; and third year growth began to steadily decrease. Growth of sockeye salmon with the same ocean history (but different broods) was highly correlated, illustrating the importance of the environment in affecting growth rates of sockeye salmon, not only in the early marine environment but later in their life history when the sockeye are more dispersed. The importance of sea surface temperature (SST), particularly during the growing season, was noted in many of the regression models. SST had its greatest influence on growth in the first ocean year where sockeye are migrating out of Bristol Bay and into the Aleutian Islands region.

Four species of juvenile salmon—pink (Oncorhynchus gorbuscha), sockeye (O. nerka), chum (O. keta) and coho (O. kisutch) salmon—were collected during July and August 1996, using a midwater trawl in near surface waters of the Gulf of Alaska from Southeast Alaska to the Alaska Peninsula. Stomach contents of these salmon were examined to identify important prey items. Crustaceans—principally hyperiid amphipods and euphausiids—and fish were the primary prey, and decapod larvae, calanoid copepods, and pteropods were also commonly found in the juvenile salmon diets. The proportions of prey type varied by habitat and regions for each of the four salmon species examined. The large variation in prey types and the small proportion, 3%, of empty stomachs suggest that the availability of prey resources does not appear to be limiting growth of juvenile salmon examined in this study.

The early marine ecology of juvenile (age -.0) Pacific salmon (Oncorhynchus spp.), their habitat utilization patterns, and annual fluctuations in abundance were studied along a seaward migration corridor in the northern region of southeastern Alaska. From May to October in 1997, 1998, and 1999, up to 24 stations spanning 250 km were sampled at approximately monthly intervals for biological and physical data in inshore, strait, and coastal habitats extended 60 km offshore into the Gulf of Alaska. Average surface (2-m) temperatures and salinities ranged from 6.9o to 13.4oC and 16.7‰ to 31.8‰. A total of 31,896 fish from 40 taxa was captured with 283 surface trawl hauls. All five species of juvenile salmon were captured and comprised 61% of the total catch; pink (O. gorbuscha) and chum salmon (O. keta) were the predominate species and comprised 29% and 24% of the catch. Catch rates of juvenile salmon were zero in May, highest in June and July, and intermediate in August and October. The highest catch rates of juvenile salmon occurred in the strait habitats in June and July. A maximum estimated aggregate density of juvenile pink and chum salmon in strait habitats in June was less than 1 fish per 1,000 m3, a density more than an order of magnitude less than the neritic carrying capacity previously reported for juvenile chum salmon. Annual differences in the apparent growth rates of juvenile pink and chum salmon appeared to be more related to differences in environmental conditions (temperature and zooplankton) than to the densities of juveniles. The seasonal decline in the abundance of juvenile salmon in strait habitats coincided with declining levels of zooplankton biomass. Catch rates of juvenile salmon in coastal habitats declined with distance offshore; most juveniles were captured over the continental shelf <25 km of shore. Juvenile salmon were eaten by 4 of the 19 fish species examined for predation, and were found in 33 (5%) of the 661 fish examined. Information on origin from marked juvenile salmon indicated stock-specific spatial and temporal habitat utilization patterns, including new documentation of Columbia River Basin stocks of stream-type chinook off Alaska in June, 3-4 months earlier that previously documented. Origins of chum salmon in the strait habitats suggest that hatchery stock groups comprise over 50% of abundance in June and July, pulse through the region coincident with the earlier component of the wild stocks, and do not appear to be impacting the carrying capacity of the region. Our results indicate that juvenile salmon have seasonal habitat utilization patterns synchronous with factors that optimize growth, however, long-term monitoring over varying environmental conditions is needed to understand relationships between early marine growth and survival, and marine habitat utilization and carrying capacity.