UBC Faculty Research and Publications

Ecosystem Change and the Decline of Marine Mammals in the Eastern Bering Sea : Testing the Ecosystem… Trites, Andrew W., 1957-; Livingston, Patricia A.; Mackinson, Steven; Vasconcellos, Marcelo, 1971-; Springer, Alan M.; Pauly, D. (Daniel) 1999

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-Trites_A_et_al_Ecosystem_change_decline.pdf [ 967.81kB ]
JSON: 52383-1.0348097.json
JSON-LD: 52383-1.0348097-ld.json
RDF/XML (Pretty): 52383-1.0348097-rdf.xml
RDF/JSON: 52383-1.0348097-rdf.json
Turtle: 52383-1.0348097-turtle.txt
N-Triples: 52383-1.0348097-rdf-ntriples.txt
Original Record: 52383-1.0348097-source.json
Full Text

Full Text

ISSN 1198-6727Fisheries CentreResearch Reports1999   Volume 7   Number 1Ecosystem Change and the Declineof Marine Mammals in theEastern Bering Sea:Testing the Ecosystem Shift andCommercial Whaling HypothesesFisheries Centre, University of British Columbia, Canada1Ecosystem Change and the Decline ofMarine Mammals in the Eastern Bering Sea:Testing the Ecosystem Shift andCommercial Whaling HypothesesbyAndrew W. Trites, Patricia A. Livingston1, Steven Mackinson,Marcelo C. Vasconcellos, Alan M. Springer 2 and Daniel PaulyFisheries Centre, University of British Columbia,  2204 Main MallVancouver, B.C. V6T 1Z4    (604) 822-81811 National Marine Fisheries Service, 7600 Sand Point Way NE, Seattle, WA 981152 Institute of Marine Science, University of Alaska, Fairbanks, Alaska 99775 1080published byThe Fisheries Centre, University of British Columbia2204 Main MallVancouver, B.C., Canada1999ISSN 1198-67272This report should be cited as follows:Trites, A.W., P.A. Livingston, M.C. Vasconcellos, S. Mackinson, A.M. Springer, and D.Pauly. 1999. Ecosystem change and the decline of marine mammals in the Eastern BeringSea: testing the ecosystem shift and commercial whaling hypotheses. Fis eries CentreResearch Reports 1999, Vol. 7 (1), 106 p.3SummarySome species in the Bering Sea underwent large changes between the 1950s and the 1980s.Among the best documented are the declines of Steller sea lions and northern fur seals, and thepossible increase and dominance of ground fish – pollock and large flatfish.  A frequentlyproposed explanation is that human exploitation of top predators and/or a shift in the physicaloceanography altered the structure of the eastern Bering Sea ecosystem.We employed two inter-related software packages (Ecopath and Ecosim) to describequantitatively the eastern Bering Sea ecosystem during the 1950s, before large-scale commercialfisheries were underway, and during the 1980s, after many marine mammal populations haddeclined.  We grouped the hundreds of species that make up the Bering Sea ecosystem into 25functional groups.Our mass-balance ecosystem models showed that most of the top predators (trophic level IV)declined from the 1950s to the 1980s.  They included Steller sea lions, seals, sperm whales,deep-water fish and other demersal fishes.  The only top predators to increase were large flatfishsuch as arrowtooth flounder.   At the mid-trophic level (III), baleen whales and pelagic fishesdeclined, while small flatfish, pollock, and walrus and bearded seals increased.  Based on ourmodel assumptions, pollock contributed over 50% of the total flow of energy at the mid trophiclevels during the 1980s compared to only 10% in the 1950s model.  In contrast, pelagic fishescontributed nearly 50% of the flow in the 1950s.  At trophic level IV, no one species dominated theflow of energy during the 1950s.  However, large flatfish contributed over 60% of the total energyflow in the 1980s model. Large flatfish and adult pollock that dominate the Bering Sea in the1980s appear to be significant competitors of seals. Large flatfish are also competitors of Stellersea lions and there are large overlaps in the diets of pollock and baleen whales.Changes in the biomass of marine mammals appear to have little effect on the biomass of othergroups in the Bering Sea.  Reductions in prey abundance can quickly reduce marine mammalpopulations, but marine mammals are unable to quickly recover when abundant food becomesavailable.   Our models suggest that Steller sea lion populations would be larger if adult pollockand large flatfish were lower in abundance due to competitive release of important prey.Most impacts on the modeled ecosystem can be associated with changing the biomass of lowertrophic levels. Total catch in the eastern Bering Sea rose from 0.33 to 2.62 t•km-2 between the1950s and the 1980s. Exploitation during the 1950s used 47% of the net primary production, withmost of it flowing through the harvested whales.  Shifting the emphasis from exploiting marinemammals in the 1950s to catching fish in the 1980s lowered the amount of primary productionrequired to sustain harvests to 6%.Some ecosystem indices derived from our ecosystem models indicate that the eastern Bering Seawas more mature in the 1950s than in the 1980s. However, we are less certain about the actualstate of the Bering Sea in the 1950s due to the relative paucity of data from that time.  Theecosystem indices for both the 1950s and 1980s models suggest that the Bering Sea is relativelyresilient and resistant to perturbations. Removing whales from the 1950s ecosystem had apositive effect on pollock by reducing competition for food.  However, whaling alone is insufficientto explain the 400% increase in pollock biomass that may have occurred between the 1950s andthe 1980s.  Nor can commercial fisheries account for these observed changes.  The magnitude ofchanges that occurred in the biomass of all the major groups in the eastern Bering Sea cannot beexplained solely through trophic interactions.  We suggest that other factors comprising a regimeshift, such as changes in water temperature or ocean currents may have been at play.4Director’s ForwardTesting the Cascade Hypothesis, Indigenous Peoples and the name of the Bering Sea.There are two reasons why the Bering Sea is mis-named.First, it already had a name. The Aleuts were one of the first human peoples to populate theislands and shores of what we call the Bering Sea and the Aleut language, which includes acreolised Russian version spoken on Bering Island, calls this sea Udaadan alagux, whichtranslates as ‘the sea around here’ or ‘our sea’, no different in kind to the Roman’s name for theMediterranean Sea, mare nostrum, which also means ‘our sea’. Using the local name for theBering Sea would be a courtesy to the indigenous Peoples of this region who have suffered notthe best of treatments as the legacy of their sea’s ‘discovery’ by Europeans.The second reason the Bering Sea is mis-named is due to a mundane mis-filing of the namegiven to it by an earlier Russian explorer (filing is sometimes said to be the secret of life). TheImperial Russian archives at Yakutsk in Siberia were something of a backwater in the 15th and16th centuries. Had they been well managed, we may now refer to the Dezhnyov Sea instead ofthe Bering Sea. In 1648 Semyon Ivanov Dezhnyov (1605-1673), a Cossack, sailed in search offurs with a fleet of seven ships around the north eastern cape of Asia and landed in Kamchatka,proving the separation of the Asian and American continents. The cape bears Dezhnyov’s name,but the records of his journey were not re-discovered until 1736 by a German historian, GerhardMüller. Unaware that the question had already been answered, in 1724 the Russian Tsar, Peterthe Great, commissioned Janasson (Vitus) Bering (1681-1741), a Danish explorer born in the 11thcentury town of Horsens, Jutland, to determine if there was a northeast passage to China. Beringdid just that, but poor weather prevented any detailed exploration, and the Bering Strait wascharted in 1730 by Mikhail Gvozdev and Ivan Fyodorov, whose trip was designed to subdue theChukchi people of the region.Soon, the Russian Empress Anna, who in 1730 emerged from the turmoil following Peter theGreat’s sudden death in 1725, commissioned Bering to return to the area again. So, when Müllerreported that Dezhnyov had been there first, a century earlier, Bering had already left St.Petersburg to explore the Siberian north, secure in the knowledge that a sea was already namedafter him. After many hardships, including building a boat in a Kamchatkan winter, Bering,captaining the St. Paul, explored much of the Alaskan peninsula and the Aleutian island chain in1741, and, by his own reports causing much distress to the aboriginal peoples who lived there.Viitus Bering and his crew, including the naturalist Georg Steller, documented the great wealth offur-bearing sea mammals in the region, exactly the news that the waiting Russian court wanted tohear. But Bering did not live to take the news home. He died of scurvy after being shipwrecked onBering Island, one of the Komandorskiye islands, famous also as the home of the Steller’s seacow, Hydrodamalis gigas, a unique cold-water sirenian and the first sea mammal to be driven toextinction by human hunting (see Pitcher 1998a). The Russian colonisation of North America hadbegun inauspiciously, like so many other similar events around the world, with the extinction ofindigenous fauna and the abrogation of native people’s rights (see Diamond 1997).Over the two centuries following the sad and final loss of the sea cow, waves of human over-exploitation of marine animals have taken place in the Bering Sea. Known by fisheries scientistsas serial depletions, in this process human hunters and fishers shift their geographical focus areaby area, and replace target species one by one. In the Bering Sea, sea otters were devastat d inthe 1700s, fur seals and coastal whales in the 1800s, and the great oceanic baleen whales in the1900s.  In each of these cases our activities have greatly altered natural ecosystems.5But what are these changes? How can we know what we have done to marine ecosystems?Until recently, there have been almost no scientific tools with which to try to answer thesequestions. Indeed, there is no end to the mischief that inadequate modelling, unable to addressecosystem linkages, can lend to such questions. The new technique of ecosystem simulationmodelling is still in its infancy, but here the ECOPATH and ECOSIM suite of techniques are used toaddress a specific hypothesis about human-caused changes in the Bering sea. ECOPATH is amass-balanced snapshot of an aquatic ecosystem that explicitly quantifies all trophic interactions,while ECOSIM allows dynamic simulations of the model biomass pools. Seven previous FisheriesCentre Research Reports have dealt with this type of ecosystem modelling: Bonfil et al. (1998),Pauly (1998), Pauly et al. (1998a), Okey and Pauly (1998), Dalsgaard and Pauly (1997), Paulyand Christensen (1996), and Pitcher (1996).In this report, these models are employed in a test of the c mp titive release hypothesis, al otermed a cascade hypothesis, as proposed in NRC (1996), that removing whales will increasenumbers of fish and other organisms that share whale food, mainly euphausiids. From anevolutionary perspective, there is no doubt that competitive interactions, and their avoidance, hasshaped the life histories and trophic niches of most species that we observe in marineecosystems. The problem is that ecologists have had great difficulty in validating such interactionsoperating in the wild. Data is always said to be insufficient, or evidence equivocal, while some averthat, without field experiments, proof of such effects is impossible (see Hall 1999). The usefulnessof the ECOPATH/ECOSIM ecosystem simulation modelling system is that these questions can beinvestigated by comparing a range of scenarios.Each ECOPATH model represents a huge amount of information culled from literature and researchwork. Models of this kind can summarise the research of many workers in constructing thedynamics of the trophic web upon which ECOSIM simulations of may be based. In the presentwork, the ECOPATH models have 26 component biomass pools. The ECOPATH/ECOSIM modellingsystem is by no means perfect, and at present is not able to capture some ecological processessuch as the structural changes caused by keystone species. But this problem can be addressedby modifying scenarios of the models to represent the past, present and future of an ecosystem, aprocess termed Back to the Future modelling (Pitcher 1998b; Haggan et al. 1998; Pitcher et al.1999). And, in addition, such models may be refined without disputing or replacing previouslyacquired information about the ecosystem, and so provide a basis for ongoing tests of a range ofhypotheses about actual changes that have been observed. Hence, this report may not be thefinal answer on the Bering Sea whale-fish interaction question, but future work may be easilybased upon what is reported here. For example, it would be valuable to try to increase precisionby incorporating the knowledge of local indigenous peoples in building models of the past. Suchknowledge is extremely valuable, especially for times before scientific data was collected, andmay have an important bearing on the questions examined in this report.The Fisheries Centre Research Reports series publishes results of research work carried out, orworkshops held, at the UBC Fisheries Centre. The series focusses on multidisciplinary problemsin fisheries management, and aims to provide a synoptic overview of the foundations, themes andprospects of current research. Fisheries Centre Research Reports are distributed to appropriateworkshop participants or project partners, and are recorded in Aquatic Sciences and FisheriesAbstracts. A full list appears on the Fisheries Centre's Web site, htpp://fisheries.com. Copies areavailable on request for a modest cost-recovery charge.Tony J. PitcherProfessor of FisheriesDirector, UBC Fisheries Centre6AcknowledgementsThis study was funded by a grant from the David and Lucile Packard Foundation.  We arealso grateful for the additional support provided by the North Pacific Marine ScienceFoundation through the North Pacific Universities Marine Mammal Research Consortium. Wewould also like to thank Pamela Rosenbaum for her administrative assistance, and PaulSpencer and Beth Sinclair of the Alaska Fisheries Science Center for providing constructivecomments on our report.7Table of ContentsSummary...................................................................................................3Director’s Forward.......................................................................................4Acknowledgements.....................................................................................6Table of Contents........................................................................................7Introduction...............................................................................................9An Overview of Ecopath & Ecosim............................................................10Eastern Bering Sea – Defining the System................................................11Species Assemblages of the Eastern Bering Sea....................................13Balancing the Ecosystem........................................................................14Balancing the 1980s model...............................................................15Balancing the 1950s model...............................................................15Ecopath Model Results............................................................................19Flow Charts & Trophic Levels............................................................22Niche Overlaps..................................................................................26Mixed Trophic Impacts......................................................................27Fisheries........................................................................................29Characterizing the Bering Sea Ecosystem............................................30Ecosim Model Results..............................................................................34Equilibrium Simulation Results............................................................36Dynamic Simulation Results..............................................................42System Recovery Time.....................................................................50Discussion of Simulation Results..........................................................51Conclusions...........................................................................................52Literature Cited........................................................................................54Appendix 1 – Mass-Balance Model Details................................................72ECOPATH - steady state mass-balance ecosystem model..........................72ECOSIM - dynamic mass-balance approach for ecosystem simulation........72Appendix 2 – Species Assemblage Details...............................................74Mammals and Birds................................................................................741.  Baleen Whales.............................................................................772.  Sperm Whales..............................................................................803.  Toothed Whales...........................................................................814.  Beaked Whales............................................................................835.  Walruses and Bearded Seals...........................................................836.  Steller Sea Lions...........................................................................847.  Seals.........................................................................................848.  Piscivorous Birds..........................................................................878Fish and Cephalopods...........................................................................879,10.   Pollock....................................................................................8911.  Deepwater Fish..........................................................................8912.  Large Flatfish..............................................................................8913.  Small Flatfish..............................................................................9014.  Pelagics......................................................................................9015.  Other Demersal fish....................................................................9116.  Cephalopods..............................................................................91Benthics and Jellies................................................................................9217.  Benthic Particulate Feeders...........................................................9218.  Infauna........................................................................................9219.  Jellyfish.......................................................................................9320.  Epifauna......................................................................................93Plankton..............................................................................................9421.  Large Zooplankton......................................................................9422.  Herbivorous Zooplankton............................................................9423.  Phytoplankton.............................................................................95Other...................................................................................................9524.  Discards and By-Catch...............................................................9525.  Detritus.......................................................................................96Appendix 3 – Parameters for the 45-Box Ecopath Model............................98Appendix 4 – Diet Matrix Tables..............................................................99Appendix 5 – Niche Overlap Tables..........................................................1039IntroductionThe Bering Sea has supported considerable indigenous and commercial demandfor fish, crustaceans and marine mammals over the past century. Despite avariety of regulations to safeguard fish, mammals and birds, some species in theBering Sea and Gulf of Alaska have undergone large and sometimes suddenpopulation fluctuations (NRC 1996). Among the best documented changes arethe declines of the Steller sea lions and northern fur seals between the 1950s and1990s, and the increase and dominance of a bottom fish – pollock and largeflatfish – since the late 1960s.A frequently proposed explanation for events over the past few decades is that acombination of changes in the physical oceanography acted in concert withhuman exploitation of top predators.  Impacts on trophic levels may havecascaded and shifted the North Pacific ecosystem to one dominated by pollock.For example, the mass removals of whales, plus the large declines of PacificOcean perch, herring, and yellowfin sole populations during the 1960s, resulted inthe rapid loss of several million tonnes of biomass.  Such an abrupt change inbiological constraints in the system may have had cascading effects that led tonew patterns of energy flow through food webs and may have affected otherspecies at other trophic levels.  These changes could have caused the recentdeclines of sea lions, harbor seals, fur seals, and certain seabirds (Merrick 1995,NRC 1996).Although a scenario of this sort is a likely explanation for events in the Bering Seaand Gulf of Alaska, no rigorous testable hypothesis has been developed. This isdue in part to gaps in our understanding of the ecosystem components, butmostly to the lack of ecosystem oriented approaches that explicitly take intoaccount trophic inter-dependencies (c.f. Laevastu and Favorite 1977).Fisheries scientists throughout the world largely agree that they must find ways toaccount for species interactions. The emerging shift of fisheries research fromsingle-species analysis towards an ecosystem-based approach requires toolsthat explicitly account for ecological interactions, especially those of a trophicnature.  Two such tools, which we employ, are Ecopath and Ecosim (Christensenand Pauly 1992a,b, 1995; Walters t al. 1997).  These are software packagesthat  explicitly describe trophic relationships between marine species and simulatechanges over time.The following describes two formal representations of the Bering Sea ecosystem– one for the 1950s, before large-scale commercial fisheries were underway, andthe second for the 1980s, after many populations of marine mammals haddeclined.  We use these models to examine how the structure of marinecommunities in the eastern Bering Sea may have changed in response todifferent fishing, harvesting and climate regimes. The models also allow us toaddress whether the physical environment acted in concert with humanexploitation of top predators to shift the ecosystem into a new domain.  We cantest whether the harvesting of whales affected the stability of the Bering Sea, or10whether the niche once occupied by whales is now filled by pollock.  We can alsoexamine how marine mammals might respond to changes in fishing pressure.We begin with a brief description of Ecopath and Ecosim, and the methods usedto compile data on the numbers, distribution, diets, and productivity of 25 groupsof marine organisms inhabiting the Bering Sea.  Detailed descriptions of eachgroup of species are contained in an adjoining appendix.  We present a flowchartshowing trophic interactions and energy flow in the eastern Bering Sea, anddescribe how the ecosystems of the 1950s differed from that of the 1980s.  Wethen explore how the ecosystem changed over time and how the structure anddynamics of fish, bird and marine mammal assemblages were affected by humanactivities.An Overview of Ecopath & EcosimThe Ecopath software is a simple approach for analyzing trophic interactions infisheries resources systems (Christensen and Pauly 1992a,b, 1995). Ecopath isbased on the earlier work of Polovina (1984), and is being widely applied toaquatic systems (Christensen and Pauly 1993, Pauly and Christensen 1995).  Itis a mass-balance approach that describes an ecosystem at steady-state for agiven period.  Further development of this steady-state model has resulted in adynamic ecosystem model called Ecosim that is capable of simulating ecosystemchanges over time (Walters t al. 1997). Ecopath and Ecosim represent all of themajor components of the ecosystem, and their feeding interactions, but arerelatively simple.  These kinds of models readily lend themselves to answeringsimple, ecosystem wide questions about the dynamics and the response of theecosystem to anthropogenic changes. Thus, they can help design policies aimedat implementing ecosystem management principles, and can provide insights intothe changes that have occurred in ecosystems over time.Ecopath models rely on the truism that:Production =biomass  accumulation + fisheries catch + mortality due topredation  + other mortality + loss to adjacent systems.This applies for any producer (e.g., a given fish population) and time (e.g., a yearor season).  Groups are linked through predators consuming prey, where: Consumption = production + non-assimilated food + respiration.The implication of these two relationships is that the system or model is mass-balanced (i.e., biomass is ‘conserved', or accounted for in the ecosystem).  Thisprinciple of mass conservation provides a rigorous framework – formalizedthrough a system of linear equations – through which the biomass and trophicfluxes among different consumer groups within an ecosystem can be estimated(Christensen and Pauly 1995; also see Appendix 1 for additional details).11Constructing an Ecopath model emphasizes ecological relationships rather thanmathematical equations.  All that is required are the types of data that areroutinely collected by fisheries scientists and marine biologists.  The model canincorporate and standardize large amounts of scattered information –  informationthat might have otherwise languished in scattered journals, reports and filingcabinets (Christensen and Pauly 1995).Ecopath is essentially a large spreadsheet that is simultaneously keeping track ofall the species and all the feeding interactions occurring within the ecosystem.  Itdescribes the ecosystem at one point in time. Ecosim, which is based on theEcopath equation, simulates how a change in one or more components mightaffect the ecosystem over time.Ecopath and Ecosim have been widely applied in recent years.  More than 80Ecopath systems have so far been published world-wide.  They span a diversityof systems including upwellings, shelves, lakes and ponds, rivers, open oceansand even terrestrial farming systems (see Christensen and Pauly 1992a,b, 1995;Walters et al. 1997; and the Ecopath home page at http://www.ecopath.org ).Eastern Bering Sea – Defining the SystemThe area we delineated in the eastern Bering Sea encompasses the regioncovered by the Alaska Fisheries Science Center’s bottom trawl surveys of theshelf and slope down to 500 m (Fig. 1).  It covers a wide range of marine habitatsthat includes shelf and slope regions, but was treated as a single homogenousarea.  Nearshore fauna and the northern portion of the Bering Sea were notconsidered. We chose these boundaries based on the availability of assessmentdata collected systematically for fisheries and marine mammals.  Total area is484,508 km2.Fig. 1.  The eastern Bering Sea as defined in the ecosystem model.  Total area isapproximately 500,000 km2.12Ecopath models were constructed for two periods: (i) the ‘1950s’ covering theperiod 1955 to 1960 and (ii) the ‘1980s’ covering the period 1979-1985.  Both areannual average models which means that the biomass, and the diets and speciescomposition of summer and winter are averaged to provide a year round ‘annualaverage’. Annual average models are better than seasonal models for examininglong-term changes because they are not overwhelmed by seasonal differences inspecies abundance and composition.  However, information on seasonalchanges is lost.Six parameters are required for each species group included in the ecosystemmodel.  They include their· biomass (the total weight of all age classes);· diet composition (the fraction of their diet made up of different species);· consumption (the total amount they eat per year);· production (accumulated biomass plus the amount that was harvested,eaten by other species, or died of other causes);· ecotrophic efficiency (the fraction of production that is exported or passedup the food web); and· export (the amount that is caught or migrates out of the ecosystem).Each species in the ecosystem model is represented by a single equationcontaining these 6 parameters.  All of the linear equations must be consistent withone another to account for all of the energy flowing in the ecosystem.  Thismeans that the Ecopath model can estimate a missing parameter for eachequation, so long as there are more equations (i.e., species groups) thanunknown parameters.  This is a simple algebraic technique (solving forunknowns) and is a useful means for deriving hard to get estimates.A couple of the parameters can be estimated in roundabout ways.  Production,for example, can be estimated from the ratio of production (P) to biomass (B).This ratio (P/B) is equivalent to the instantaneous rate of total mortality (Z) usedby fisheries biologists (Allen 1971).  Similarly, consumption (Q) can be expressedas a ratio of consumption to biomass, i.e., Q/B.  Dividing P/B by Q/B gives P/Q.This ratio (P/Q) is a measure of gross food conversion efficiency (i.e., the fractionof energy that a species converts into production).  In most cases, efficiencyranges from 0.1 to 0.3, which means that most groups consume 3-10 times morethan they produce.13Species Assemblages of the Eastern Bering SeaThere are hundreds of species n the Bering Sea ranging from tiny phytoplankton,through invertebrates, fish, birds and mammals.  Many of the species inhabit theBering Sea year round while others are seasonal migrants.  Creating a separatebox in our ecosystem flow chart for each of the species would require anenormous amount of information and detail.  Such detail may not enhance ourunderstanding of the ecosystem, and may detract from gaining insight into thecomponents that interest us most.  Fortunately many species share functionalsimilarity in an ecosystem and can be aggregated into ‘functional groups’, eachrepresented by an Ecopath box.We began by considering 45 groupings of species in the eastern Bering Seaecosystem. Using data (see following) on biomass, production, consumption,trophic level and diet composition, we used the ‘user-controlled aggregation’option in Ecopath to further aggregate these 45 groups.  We selected the groupsto be aggregated and did not use the automatic aggregation option. We soughtsimilarities in biological rates (P/B and Q/B), taxa and diet composition, andarrived at the following 25 boxes:Mammals and Birds1. Baleen whales - Fin, Minke, Blue, Humpback, Bowhead, Right, Gray;2. Sperm whales;3. Toothed whales - Beluga, Killer, Dalls porpoise, Harbour porpoise;4. Beaked whales - Stejnegers spp.;5. Pacific Walrus and Bearded Seals;6. Steller sea lions;7. Seals - Northern fur, Harbour seal, Spotted seal, Ribbon seal, Ringed seal8. Piscivorous Birds - murres, kittiwakes.Fish and Cephalopods9. Adult pollock;10. Juvenile pollock;11. Deepwater fish - Sablefish, rockfish, Macrouridae (rattails);12. Large flatfish - Halibut, Greenland turbot and arrowtooth flounder;13. Small flatfish - Yellowfin sole, rock sole, Alaska plaice and flathead sole;14. Pelagics - Herring and other pelagic fish (lanternfish, capelin, sandlance,salmon);15. Other demersal fish - Cod, skates, sculpins, and eelpouts;16. Cephalopods - Squid, octopus.Benthics and Jellies17. Benthic Particulate Feeders. – Tanner crab, King crab and shrimp;18. Infauna - Clams, polychaetes, other worms;19. Jellyfish;20. Epifauna - Hermit crabs, snails, brittlestars, starfish.14Plankton21. Large Zooplankton - Euphausiids, Mysids, amphipods;22. Herbivorous Zooplankton – Calanoids;23. Phytoplankton.Other24. Discards;25. Detritus.Summary parameters for the species and aggregated groups are contained inTables 1 and 2. Choice of parameters and critical assumptions are outlined inAppendix 2. Information gathered from published sources included the numbersor biomass of each group of species living in the Bering Sea; their diets, rates ofconsumption and production. The two primary data sources for groundfishbiomass estimates were surveys or stock assessments conducted by the SovietUnion during the 1950s and the United States National Marine Fisheries Service(NMFS) during the 1980s (see Appendix 2).  Population estimates for marinemammals came from Perez (1990), the NRC Report (1996), and the 1987-1988Marine Mammal Protection Act annual report (NMFS 1988).  Estimates for otherspecies were drawn largely from stock assessments performed by NMFSscientists.  Diet data for groundfish comes primarily from the groundfish foodhabits data base of NMFS, while diet data for marine mammals comes primarilyfrom the 1996 NRC Report and references therein.  Several assumptions weremade with respect to the 1950’s model.  For groups that had no abundanceinformation from that time period, we assumed they had the same biomass as the1980’s period.  We also made assumptions about the diet compositions ofanimals during that time period, particularly assuming that pelagic fish (herring,capelin, and others) were a more important prey than pollock in the 1950s.Balancing the EcosystemWe compiled a large number of parameters to estimate the flows of energybetween the 45 boxes we initially chose to describe the eastern Bering Seaecosystem in the 1980s (Appendix 3).  Using the ‘controlled aggregation’ option inEcopath, these 45 species groupings were re-grouped into the 25 boxesdescribed above.  However, it is unclear at this point, whether the estimates andassumptions we have outlined are reasonable and mutually compatible.  One testof our choice of parameters is to determine whether they jointly lead to abalanced ecosystem (i.e., one that is at equilibrium where inputs into the boxesare balanced by the output of these boxes).  If our description of the easternBering Sea ecosystem balances, there will be sufficient prey to support the foodrequirements of the predators, and the energy inputs and outputs will balanceeach other.15A number of approaches can be taken to balance our 1950s and 1980s models.In our case, we estimated the ‘ecotrophic efficiencies’ (EE) for each of the 25groups of species.  Ecotrophic efficiency is the fraction of production that isexported or passed up the food web.  It cannot exceed 1.0 under the equilibriumassumption.  An ecotrophic efficiency greater than 1.0 identifies a group that isnot receiving sufficient input from lower parts of the food web.  It indicates to usthat some of the initial parameter estimates may have to be verified or modified.A value less than 1.0 means that more energy is entering than is exiting, with thedifference ending up as accumulated detritus or being exported from theecosystem.Balancing the 1980s modelOur initial estimates for the 25 species groups were consistent with one another,with four exceptions (Table 1).  The changes needed to balance the modelincluded:· Increasing the phytoplankton biomass from 21 to 32 t•km-2;· Increasing the Q/B ratio of jellyfish from 1.09 to 2.00 year-1 to nsure P/B >(Q/B)/2;· Reducing the harvest of walrus and bearded seals by half. (The highecotrophic efficiency of 2.076 meant that there were either too many sealsand walruses harvested or we had underestimated the actual numberpresent.); and· Reducing the proportion of discards eaten by other species.  (This assumesthat the groups of species were able to compensate for the reduction in dietcoming from discards by consuming comparable increases of their major preygroups).Balancing the 1950s modelThe most important assumptions we had to evoke to balance the 1950s model(Table 2) concerned diet compositions. The most notable was our assumptionthat pelagic fish (herring, capelin, and others) were a more important prey thanpollock in the 1950s.  We assumed that pollock were in low abundance during the1950s, based on information from non-systematic Soviet trawl surveys during thatperiod.  Hence, pelagic fish likely comprised the major part of many predatordiets.16Table 1. Ecopath parameters describing the 1980s eastern Bering Sea ecosystemwith 25 functional groups where P/B is the ratio of production to biomass and Q/Bis the ratio of consumption to biomass.Functional Biomass P/B Q/B CatchGroup (t km-2) (year-1) (year-1) (t km-2)1.  Baleen whales 0.394 0.020 11.383 0.0002.  Toothed whales 0.009 0.020 13.108 0.0003.  Sperm whales 0.208 0.020 4.553 0.0004.  Beaked whales 0.001 0.020 10.515 0.0005.  Walrus & bearded 0.074 0.060 11.249 0.0046.  Seals 0.066 0.060 15.926 0.0017.  Steller sea lions 0.019 0.060 12.702 0.0018.  Pisc. Birds 0.006 0.800 60.000 0.0009.  Adult pollock 27.451 0.500 2.640 1.89510. Juvenile Pollock 6.000 2.500 8.333 0.00011. Other demersal fish 3.904 0.433 2.226 0.12812. Large flatfish 1.900 0.400 2.444 0.05013. Small flatfish 9.161 0.400 2.968 0.21114. Pelagics 13.644 0.798 3.650 0.21215. Deepwater fish 0.407 0.400 2.490 0.00716. Jellyfish 0.048 0.875 2.000 0.00017. Cephalopods 3.500 3.200 10.667 0.00018. Benth.Par. feeders 5.800 1.480 7.690 0.10819. Infauna 46.500 1.373 12.000 0.00020. Epifauna 5.858 1.578 5.777 0.00021. Large Zoops 44.000 5.091 22.000 0.00022. Herb. Zoops 55.000 6.000 22.000 0.00023. Phytoplankton 32.000 60.000 0.000 0.00024. Discards 0.000 - - 0.00025. Detritus 0.000 - - 0.00017Table 2.  Ecopath parameters describing the 1950s eastern Bering Sea ecosystemwith 25 functional groups where P/B is the ratio of production to biomass and Q/Bis the ratio of consumption to biomass.Functional Biomass P/B Q/B CatchGroup (t km-2) (year-1) (year-1) (t km-2)1.  Baleen whales 0.696 0.020 13.678 0.0842.  Toothed whales 0.009 0.020 13.108 0.0003.  Sperm whales 0.439 0.020 4.553 0.0214.  Beaked whales 0.001 0.020 10.515 0.0005.  Walrus & bearded 0.054 0.060 11.651 0.0066.  Seals 0.106 0.060 15.577 0.0057.  Steller sea lions 0.029 0.060 12.703 0.0018.  Pisc. Birds 0.006 0.800 60.000 0.0009.  Adult pollock 5.500 0.500 2.640 0.01410. Juvenile Pollock 0.942 2.500 8.333 0.00011. Other demersal fish 8.957 0.433 2.226 0.00112. Large flatfish 1.169 0.400 2.444 0.00213. Small flatfish 8.530 0.400 2.968 0.10514. Pelagics 28.869 0.798 3.650 0.08315. Deepwater fish 1.011 0.400 2.490 0.00116. Jellyfish 0.048 0.875 2.000 0.00017. Cephalopods 3.500 3.200 10.667 0.00018. Benth.Par. feeders 29.000 1.480 7.690 0.01019. Infauna 75.000 1.373 12.000 0.00020. Epifauna 8.000 1.578 5.777 0.00021. Large Zoops 44.000 5.091 22.000 0.00022. Herb. Zoops 55.000 6.000 0.00023. Phytoplankton 32.000 60.000 0.00024. Discards 0.000 - - 0.00025. Detritus - -18Our assumption that species consumed more pelagic fish than pollock in the1950s resulted in nine groups having and ecotrophic efficiency above one:· Baleen whales EE=6.051;· Sperm whales EE=2.392;· Walrus and Bearded EE=1.888;· Phytoplankton EE=1.410;· Adult pollock EE=1.415;· Small flatfish EE=1.042;· Cephalopods EE=1.052;· Infauna EE=1.334; and· Epifauna EE=1.352.Ecotrophic efficiencies above 1.0 indicate that demand upon them is too high tobe sustainable.  We know that this was indeed the case for whales (and perhapswalruses).  Harvest rates would have had to be 50-85% less to be sustainable, orthe populations of walruses and whales would have had to be 2-6 times larger tosupport the numbers being removed each year (Table 3).  Either possibility, or acombination of them, would have reduced the ecotrophic efficiency to 1.0 or less.Table 3. Reduction in harvest of walrus, bearded seals, and sperm and baleenwhales that would balance production.  The estimates were calculated by assuminga consumption of zero and no export from the system other than harvest (Underthese conditions the Ecopath master equation can be reduced to B•[P/B] •EE-H=0,where EE=H/P and H=P when EE=1).Group Current Modified(t km -2) (t km -2) % factorBaleen Whales 0.084 0.0139 83 6.0 x lessSperm Whales 0.021 0.0088 58 2.4 x lessWalrus & Bearded 0.006 0.003 50 2.0 x lessChangeWe changed demands on the other groups with high ecotrophic efficiencies asfollows:· Phytoplankton biomass was increased to 32 t•km-2 as detailed above;· We slightly adjusted the relative importance of small flatfish and cephalopodsin the diets of other groups;· Biomass of pollock was increased from 4.32 to 5.50 t•km-2 (reducing theimportance of pollock in the diet of its predators reduced the EE by very little);and· Infauna was increased from 55 to 75 t•km-2 and epifauna from 5.8 to 8 t•km-2,since adjusting the importance of infauna and epifauna in the diets of theirpredators had a minimal effect on EE.The only other change required to balance the 1950s ecosystem was to increasethe Q/B ratio for jellyfish from 1.09 to 2.00 year-1.19Ecopath Model ResultsA flowchart showing trophic interactions and energy flow in the eastern BeringSea during the 1980s is presented in Fig. 2. It shows the estimated trophic levelof each of our 25 functional groups and the relative amounts of energy that flow inand out of each box. Large numbers of flows in the Bering Sea emanate fromthree species at trophic level III–pollock, small flatfish and pelagic fishes.  Majorconsumers are the top–trophic level IV–predators.  They include the marinemammals and birds, as well as large flatfish and deepwater fish.Fig. 2.  Flowchart of trophic interactions in the eastern Bering Sea during the 1980s.  All flowsare in t•km-2 year-1.  Minor flows are omitted as are all backflows to the detritus.  The size ofeach box is roughly proportional to the biomass therein.20Basic parameters for the 1950s and 1980s ecosystem models are contained inTables 1 and 2.  Changes in biomass between the two modelled time periods areshown in Table 4.   Diet compositions are in Appendix 4.Table 4. Comparison of changes in biomass from the 1950s to the 1980s models.Functional ChangeGroup 1950s 1980s (%)1.  Baleen whales 0.696 0.394 -432.  Toothed whales 0.009 0.009 03.  Sperm whales 0.439 0.208 -534.  Beaked whales 0.001 0.001 05.  Walrus & bearded 0.054 0.074 376.  Seals 0.106 0.066 -387.  Steller sea lions 0.029 0.019 -348.  Pisc. Birds 0.006 0.006 09.  Adult pollock 5.500 27.451 39910. Juvenile Pollock 0.942 6.000 53711. Other demersal fish 8.957 3.904 -5612. Large flatfish 1.169 1.900 6313. Small flatfish 8.530 9.161 714. Pelagics 28.869 13.644 -5315. Deepwater fish 1.011 0.407 -6016. Jellyfish 0.048 0.048 017. Cephalopods 3.500 3.500 018. Benth.Par. feeders 29.000 5.800 -8019. Infauna 75.000 46.500 -3820. Epifauna 8.000 5.858 -2721. Large Zoops 44.000 44.000 022. Herb. Zoops 55.000 55.000 023. Phytoplankton 32.000 32.000 024. Discards 0.000 0.000 -25. Detritus - - -Biomass (t km -2)21Various ecosystem indices that relate to resilience, resistance to perturbations,competition between species, and flows between ecosystem elements are shownfor the 1950s and 1980s ecosystems in Table 5.  These indices, drawn fromtheoretical ecology (Odum 1969, Holling 1973, Christensen 1995), allow theecological characteristics of the eastern Bering Sea to be compared over time(1950s-1980s) and with other marine ecosystems (Vasconcellos et al. 1997,Christensen and Pauly 1998).   Among the properties of ecosystem models thatcan be determined are niche overlaps, biomass pyramids, relative contribution ofspecies to the total flow of energy in the ecosystem and the amount of primaryproduction required to sustain marine mammals, fish and fisheries.Table 5. Descriptive summary statistics for the 1950s and 1980s eastern Bering Seaecosystem models.Descriptive Indices 1950s 1980s UnitsSum of all consumption 3576.11 3073.72 t km-2•year-1Sum of all exports 0.33 2.62 t km-2•year-2Sum of all respiratory flows 1885.15 1620.43 t km-2•year-3Sum of all flows into detritus 1073.37 994.99 t km-2•year-4Total system throughput 6534.97 5691.76 t km-2•year-5Sum of all production 2679.77 2612.84 t km-2•year-6Mean trophic level of fishery catches 3.44 3.30 -Gross efficiency (catch/net p.p.) 0.0002 0.0021 -Input total net primary production 1770 1264 t km-2•year-6Calculated total net primary production 1920 1920 t km-2•year-6Unaccounted primary production - - -Total primary production/total respiration 0.94 0.78 -Net system production 115.15 356.43 t km-2•year-6Total primary production/total biomass 5.85 4.94 -Total biomass/total throughput 0.046 0.045 -Total biomass (excluding detritus) 302.44 255.95 t km-2•year-6Total catches 0.33 2.62 t km-2•year-6Connectance Index 0.29 0.30 -System Omnivory Index 0.183 0.157 -22Flow Charts & Trophic LevelsThe flow chart depicting the 1950s ecosystem is similar in layout to the 1980sflowchart.  It contains the same species at roughly the same trophic levels (Table6).  Where the two systems differ is in the relative sizes of the boxes (i.e., in thebiomass of the different functional groups – Figs. 3 and 4).Groups near trophic level IV that were lower in abundance in the 1980s relative tothe 1950s included seals, Steller sea lions, sperm whales, deep water fish andother demersal fishes.  The only group that was higher in abundance was largeflatfish such as arrowtooth flounder.  At the next level down (near trophic level III)pollock, small flatfish and walruses were estimated to have higher biomass in the1980s relative to the 1950s, while pelagic fishes, such as herring and sandlance,were in low abundance in the 1980s.  The 1980s biomass of benthic particulatefeeders (crabs), epifauna and infauna was also low.Fig. 3.  Flowchart of trophic interactions in the eastern Bering Sea during the 1980s.The blackened boxes indicate which groups had lower estimated abundance in the1980s than in the 1950s, and the shaded boxes show which species were estimatedto have higher abundance in the 1980s than in the 1950s.  Connecting lines showthe major trophic flows of energy between functional groups (minor flows areomitted).23Table 6.  Estimated trophic levels of the 25 groups of species in the 1950s and1980s models.1950s 1980s 1950s 1980sSperm 4.75 4.71 Adult pollock 3.34 3.29Beaked whales 4.59 4.58 Pelagics 3.25 3.20Toothed whales 4.37 4.33 Small flatfish 3.20 3.16Steller sea lion 4.30 4.24 Jellyfish 3.16 3.15Pisc. birds 4.13 4.03 Juvenile pollock 3.09 3.07Large flatfish 4.12 4.02 Benth.P. feeders 2.82 2.78Deepwater fish 4.01 4.05 Epifauna 2.63 2.38Seals 4.01 3.95 Large Zooplankton 2.28 2.22O. demer. fish 3.82 3.85 Infauna 2.00 2.00Cephalopods 3.76 3.71 Herb. Zooplankton 2.00 2.00Baleen whales 3.61 3.60 Phytoplankton 1.00 1.00Walrus& Bearded 3.51 3.53 Detritus 1.00 1.00Fig. 4.  Estimated trophic levels and relative abundance of species in the easternBering Sea during the 1980s.  Blackened boxes indicate groups that had lowerabundance in the 1980s relative to the 1950s, and shaded boxes show species thathad higher abundance in the 1980s relative to the 1950s.  Major flows of energybetween the boxes are shown in Figs. 2 and 3.24The most important differences between the 1950s and 1980s eastern BeringSea ecosystem models occurred at the high trophic levels, characteristicallyoccupied by large fish and marine mammals.  The relative contribution of thedifferent functional groups to the total flow of energy (expressed in Ecopath as‘throughput’ which equals the sum of all flows from consumption, export,respiration, and detritus) was calculated for trophic levels III and IV.  At level III(Fig. 5), adult and juvenile pollock contributed almost 50% of the total flow ofenergy during the 1980s compared to less than 10% during the 1950s.  Therelative importance of species occupying trophic level III shifted from pelagicspecies in the 1950s model to pollock in the 1980s model.TROPHIC LEVEL III1950s0 20 40 60 80PelagicsCephalopodsSmall FlatfishO.Dem.FishAdult pollockBaleen whalesJuv. pollockWalrus&BeardedJellyfish% flow1980s0 20 40 60 80Adult pollockJuv. pollockPelagicsCephalopodsSmall FlatfishO.Dem.FishBaleen whalesSealsWalrus&BeardedJellyfish% flowFig. 5.  Relative contribution of species (in %) to the total flow of energy(throughput) at trophic level III in the 1950s and 1980s models.  Note that the sum ofall bars in each panel equals 100%.25Changes at trophic level IV were mainly due to the increase in the dominance oflarge flatfishes in the total throughput (Fig. 6).  During the 1950s, the flow ofenergy was much more evenly distributed among species due to the largebiomass of seals and sperm whales.  Seals fed at a slightly higher trophic levelduring the 1950s (4.01) compared to the 1980s (3.95, Table 6).Detritus is an important component of the eastern Bering Sea ecosystem.  Totalflow originating from detritus rose from 26% in the 1950s to 34% in the 1980s.This rise in throughput from the detritus indicates more groups of species feedingon detritus.  Usually a high utilization of detritus indicates a mature (Christensenand  Pauly  1998)  and  resilient  ecosystem   (Vasconcellos  et  al.  1997).     OurTROPHIC LEVEL IV1950s0 20 40 60 80Large FlatfishDeepwater fishSperm whalesSealsSteller SealionPisc. birdsToothed whalesBeaked whales% flow1980s0 20 40 60 80Large FlatfishDeepwater fishSperm whalesPisc. birdsSteller SealionToothed whalesBeaked whales% flowFigure 6. Relative contribution of species to the total flow of energy (throughput) attrophic level IV in the 1950s and 1980s models.  Note that the sum of all bars ineach panel equals 100%.26interpretation is slightly complicated by our treatment of discards as detritus in the1980s model.   However, calculations by Queirolo et al. (1995) estimate thatdiscards are less than one percent of the total detrital pool in the eastern BeringSea, an indication that the detrital throughput in the 1980's model is a sign of amore mature benthic ecosystem,Niche OverlapsNiche overlaps were calculated between pollock, large flatfish and marinemammals in the 1980's model (Table 7).  We took two approaches.  One was todetermine to what extent any two groups seek the same prey (referred to as preyoverlap).  The other was to determine to what extent they are subject to predationby the same predators (predator overlap).  A value of 1 indicates completeoverlap in the resources they share (prey overlap), or in the predators thatconsume them (predator overlap). A value of 0 indicates that the groups do notshare resources or are not preyed upon by the same predators. Both indices areimportant and can be used to describe various kinds of niche partitioning betweengroups.Table 7. Estimated niche overlaps for marine mammals, pollock and large flatfish inthe 1980's model.  See text for explanation.  Dashes indicate no overlap, and valuesof 0.00 indicate an overlap <0.01.  Overlaps ³ 0.50 are in bold characters.Prey OverlapGroup Name 1 2 3 4 5 6 7 8 9 10   1. Baleen whales 1.00 - - - - - - - - -   2. Toothed whales 0.48 1.00 - - - - - - - -   3. Sperm whales 0.30 0.51 1.00 - - - - - - -   4. Beaked whales 0.32 0.74 0.61 1.00 - - - - - -   5. Walrus&bearded 0.15 0.29 0.01 0.21 1.00 - - - - -   6. Seals 0.54 0.66 0.04 0.38 0.39 1.00 - - - -   7. Steller sea lion 0.24 0.49 0.15 0.26 0.09 0.40 1.00 - - -   8. Adult pollock2+ 0.86 0.16 0.00 0.04 0.06 0.50 0.08 1.00 - -  9. Juv. pollock0-1 0.73 0.04 - - - 0.21 - 0.86 1.00 - 10. Large flatfish 0.26 0.35 0.03 0.15 0.11 0.45 0.38 0.34 0.08 1.00Predator OverlapGroup Name 1 2 3 4 5 6 7 8 9 10   1. Baleen whales 1.00 - - - - - - - - -   2. Toothed whales 1.00 1.00 - - - - - - - -   3. Sperm whales - - - - - - - - - -   4. Beaked whales - - - - - - - - - -   5. Walrus&bearded 1.00 1.00 - - 1.00 - - - - -   6. Seals 1.00 1.00 - - 1.00 1.00 - - - -   7. Steller sea lion - - - - - - - - - -   8. Adult pollock2+ 0.00 0.00 - - 0.00 0.00 - 1.00 - -   9. Juv. pollock0-1 - - - - - - - 0.15 1.00 - 10. Large flatfish 0.05 0.05 - - 0.05 0.05 - 0.03 0.71 1.0027In terms of who is seeking the same prey in the Bering Sea, baleen whales andpollock (both adult and juvenile) have the greatest dietary overlap 73-86% (nicheprey –Table 7). There is also a significant amount of competition between sealsand adult pollock, and between adult pollock and pelagic fishes (Appendix 5).Toothed whales compete primarily with beaked whales and seals, while adultpollock share a large proportion of their diet with juvenile pollock. The largestcompetitors of sea lions appear to be seals, toothed whales and large flatfish(Table 7).  Niche overlaps for other species in our model are contained inAppendix 5.Toothed whales consume adult pollock and other species of whales, but theniche overlap is small (niche predator – Table 7). Similarly, adult and juvenilepollock only have a small overlap in their predators (15%).  The fish groups mostoften targeted by the same predators are juvenile pollock and large flatfish (71%).Mixed Trophic ImpactsWe used a Leontif matrix (see Christensen and Pauly 1992b) to explore the directand indirect impacts of competition and predation on species in the Bering Sea.The matrix assesses how an increase in the biomass of one group affects thebiomass of another (Figure 7). Impacts shown in the figure are relative butcomparable between groups. The Leontif matrix shows how changes at one levelof the food web affect others (cascade effects) and can be regarded as a form ofsensitivity analysis (i.e., how sensitive groups are to changes in the biomass ofanother group).  It can also give some insight into the stability of the ecosystem interms of its ability to withstand changes.The mixed trophic impact graph indicates that a change in the biomass of marinemammals has little or no effect on changes in the biomass of other groups.Another obvious feature is that most impacts on the ecosystem are associatedwith changing the biomass of lower trophic levels.  Increasing the biomass at alow or mid trophic level has a positive effect on higher trophic levels (presumablymore food is available) and has a negative effect on other low trophic levels (foodis presumably reduced through competition).Increasing the biomass of adult pollock has a positive effect on the fishery and anegative effect on juvenile pollock.  Adult pollock cannibalize juvenile pollock andout-compete the seabirds feeding on juvenile pollock (note however that thematrix does not account for an increase in juvenile abundance that should occurthrough increased recruitment).Our models assumed that pelagic fishes comprised over 60% of the diet of sealions in the 1980s and over 80% of the diet in the 1950s.  Increasing the pelagicfish population in the models results in increases in the Steller sea lionpopulations.  Pelagic fishes in turn are possibly influenced by the amount ofphytoplankton and large zooplankton.  Increases in plankton not only benefitSteller sea lions, but  they  also  have  positive  effects  on  all  of the upper trophic281950s1980sFig. 7. The Leontif matrix showing mixed trophic impacts in the eastern Bering Seaecosystem in the 1950s and 1980s models.  Increasing the abundance of specieson the Y-axis has positive (black bar), negative (grey bar) or no effect on specieslisted on the X-axis.  Impacts are expressed as percent changes, and are relativeand comparable between groups.29levels – most notably the birds and mammals.  Greater primary productionincreases overall biomass.Combining the information from predator and prey niche overlaps, and the mixedtrophic impacts can generate hypotheses regarding changes in the structure ofthe ecosystem. These are tested later using EcosimFisheriesWe estimated the total catch in the eastern Bering Sea was 0.33 t•km-2 in he1950s and 2.62 t•km-2 in the 1980s.  Fisheries for whales, seals and fish were thesole exports from our modeled Bering Sea ecosystem.The ratio of catch to primary production is a measure of gross efficiency of thefisheries.  Higher ratios are expected for fisheries harvesting lower in the foodchain.  Lower values indicate a fishery that specializes on apex predators(Christensen and Pauly 1993) or mid-trophic level marine mammals such asbaleen whales which have high consumption but low production rates.  Mostfisheries in the world have a weighted average ratio of 0.0002.The ratio of catch to primary production in the eastern Bering Sea was 0.0002 inthe 1950s and 0.0021 in the 1980s (Table 4).  This 10-fold increase reflects theshift from primarily harvesting mostly mid-trophic level baleen whales and pelagicfish   to catching primarily pollock, a mid-trophic level fish.  The shift  seen in themean trophic level of the catch, which moved from 3.44 (1950s) to 3.30 (1980s),is a reflection of this shift from baleen whales to pollock. Although baleen whalesconsume mostly zooplankton, the slightly higher trophic level estimates of baleenwhales relative to pollock is due to the  the presence of squid, a high trophic levelspecies, in the baleen whale diet.  Thus, the declining trend in trophic level of thecatch is not due to a fishing down the food web effect that has been noted in otherexploited marine ecosystems (Pauly et al. 1998b), but rather is due to a switch intargeting between two mid-trophic level species with slightly different diets.  Iftrophic level of the catch in the Bering Sea is estimated using only fish andinvertebrate catch and excluding the whaling of the 1950s and 1960s, there hasactually been an increase in the estimated trophic level of the catch in the easternBering Sea from the 1950s to the present (Queirolo et al., 1995).  The ratio ofcatch to primary production during the 1980s is also comparable to other shelfsystems around the world where fishing concentrates on mid-trophic level fishspecies such as sardines and anchovies, as opposed to the 1950s Bering Seasystem in which fishing was focused on mid-trophic level marine mammals.Nearly 40% of net primary production in terrestrial systems is used to sustainagriculture, industry and other activities (Vitousek et al. 1986).  Approximately 8%is required in aquatic systems (Pauly and Christensen 1995). However, this figureconceals a huge difference between coastal shelves where 25-35% of primaryproduction is required to sustain 90% of the world’s fisheries that is taken fromcoastal shelves, and the large open ocean which sustains relatively smallfisheries.  Notable exceptions are Pacific salmon which rely heavily on the30production of the open North Pacific Ocean, but are caught nearshore or in freshwater habitats.  We estimated that the fisheries operating during the 1950srequired 47% of the net primary production of the eastern Bering Sea, with mostof it flowing through the whales that were removed by commercial whalers.  Theshifting emphasis from exploiting marine mammals (which have highconsumption and low production rates) in the 1950s to catching fish (which havelower consumption and higher production rates) in the 1980s has lowered theneed for primary production to 6.1%.  About half of this primary production isrequired to sustain the pollock fishery.Characterizing the Bering Sea EcosystemTable 5 contains many indices that characterize the eastern Bering Sea.  All told,they indicate an ecosystem that according to some measures was more maturein the 1950s than in the 1980s.  They also suggest a system in both time periodsthat is relatively resilient and resistant to perturbations. The indices allow theecological characteristics of the eastern Bering Sea to be compared over time(1950s-1980s) and with other marine ecosystems.Biomass pyramids are particularly useful for inter-system comparisons (Fig. 8).The volume of each compartment of a pyramid is proportional to its throughput,and  the  top  angle of the  pyramid  is  inversely  proportional to the mean  trophicTable 8. Estimated net primary production required (PPR) to sustain fisheries forwhales, seals and fish during the 1950s and 1980s.  PPR is expressed as apercentage of total primary production available in the ecosystem.1950s 1980sBaleen whales 33.7 -Sperm 9.7 -Walrus & Bearded seal1.0 0.7Seals 2.2 0.4Steller Sea lions 0.3 0.2Adult pollock 0.0 2.8O. demer. fish 0.0 0.7Large flatfish 0.0 0.2Small flatfish 0.3 0.6Pelagics 0.1 0.2Deepwater fish 0.0 0.1Benth.P. feeders 0.0 0.1Total 47.3 6.1PPR (%)31transfer efficiencies at trophic levels II-IV. The pyramids show how the energy-flows between groups are distributed between the different trophic levels.Only the trophic pyramids for the 1950s model are shown in Fig. 8 because theyare very similar to the 1980s pyramid.  Both pyramids indicate that biomass andenergy flow are distributed fairly well throughout system.  They also show thatapex predators at trophic level IV do not contribute much to the biomass of theeastern Bering Sea. Unlike biomass, flows (Fig. 8) are largely contained within thefirst three trophic levels of the system. The steep sided pyramid indicates thatthere is a lot of flow within these lower trophic levels.The trophic pyramids (Fig. 8) and ecosystem indices (Tables 5 and 9) indicatethat the eastern Bering Sea is a mature system compared to other shelf systems.However, they also show that in some ways the Bering Sea was less mature inthe 1980s model compared to its state in the 1950s model.The ratio of primary production to respiration reflects the maturity anddevelopment of an ecosystem.  This ratio was close to 1.0 in the 1950s model(Table 5) and suggests a mature system where the amount of energy that wasfixed in the ecosystem is balanced by the cost of maintenance.  The value in the1980s model was lower (0.74), indicating an ecosystem that is less mature.  Thelower value in the 1980s model primarily reflects our assumptions of lowerbiomass (and thus respiration) of benthic infauna, which provides a largeproportion of the contribution to total system respiration, in the 1980s versus the1950s models.  Benthic infaunal surveys during the two periods used differentsurvey methods and may not be comparable.  Therefore, conclusions aboutsystem maturity based on our knowledge of benthic infauna biomass are possiblypremature.  Similarly, our conclusions about changes in system maturity basedon the net amount of system production in the two periods and by the ratio of totalbiomass to energy throughput (Table 5), which are also based on the changes inrespiration values due to our assumptions about benthic infauna biomass,  shouldalso be tempered by our uncertainty about these values in the two time periods.Large values of net system production (the difference between total primaryproduction and respiration) are expected in an immature ecosystem, and valuesclose to zero are expected in mature ones. The estimated value  is more than twotimes greater in the 1980s than the 1950s model, suggesting a loss in maturity.As noted above, the main reason for the higher value in the 1980s is due to thelarger contribution of benthic infauna to the total respiration in the 1950s becauseof our assumption of higher benthic infauna biomass during that time period,which is a relatively uncertain assumption.Another index that also suggests that the ecosystem was more mature in the1950s is the ratio of biomass to throughput, which was slightly higher in the 1950sthan in the 1980s (Table 5). This index expresses the proportion of biomass thatcan be supported by the available flow in a system; and will rise in value as asystem matures.  Again, the changes in this index reflect the changes inrespiration and biomass in the system that are primarily due to our assumptionsabout benthic infauna biomass in the 1950s relative to the 1980s.  Furtherassessment of the level of benthic infauna biomass during the present time mayhelp us better understand the biomass dynamics of this group over time.32Fig. 8.  Trophic pyramids representing the distribution of biomass and energy flowin four ecosystems.  The pyramids are scaled so that the volume at each trophiclevel corresponds to the sum of all flows at that level.  The top angles are inverselyproportional to the transfer efficiency (acute angle = high efficiency).One index that counters our conclusion about the loss of maturity is the ratio ofprimary production to total biomass (Table 5).  In theory, biomass shouldaccumulate in an immature system, and the ratio of primary production tobiomass should decline as the system matures.  Thus, we expected the value forthe 1980s (4.94) to exceed the value for the 1950s (5.85). A declining ratio withincreasing maturity would be expected under an assumption of similar primaryproduction over the two periods.  However, as noted in Appendix 1, we assumedlower primary production during the 1980s based on baleen isotope data fromanimals that were believed to feed primarily in the northern Bering Sea.   But thisassumption is not consistent with chlorophyll concentrations (another indicator ofprimary production) reported by Sugimoto and Tadokoro (1997).  They indicatethat primary production increased in the southeastern Bering Sea (the main areaencompassed by our model) from the 1950s to the 1980s.  A more completeunderstanding of the time trends in primary production in the North Pacific isneeded.     Furthermore, given the possibility for non-constant primary production33Table 9. Comparative statistics for the eastern Bering Sea models and other shelfmodels.Ecosytems Through Catch PP/B B/T Net syst.OmnivoryAscen- Cycling Path-put PP prod. Index dency index lengthYacutan 2362 0.0029 27.4 0.036 370 0.134 44.0 2.8 2.84N. Gulf of Mexico 1790 0.0002 7.0 0.015 19 0.195 39.1 2.1 3.03Venezuela (upwl.) 5309 0.0016 27.0 0.023 831 0.135 39.9 2.2 4.05Brunei, SE Asia 1816 0.0008 28.6 0.018 300 0.201 29.4 16.3 2.80Peru 70 (upwell.) 18800 0.0017 87.5 0.012 14709 0.169 38.1 8.7 3.63Monterey 17513 0.0012 1.2 0.012 2208 0.324 66.2 4.4 3.63Alaska Gyre 5946 - 38.1 0.015 407 0.103 42.3 - 2.03BC Shelf 1237 - 21.1 0.180 4106 0.140 40.1 - 2.03Bering S. 50’s 6535 0.0002 5.9 0.050 -115 0.183 32.5 13.2 3.47Bering S. 80’s 5692 0.0021 4.9 0.050 -356 0.157 30.9 11.1 3.51over time, we may need to derive new system maturity indices that do not rely onan assumption of constant primary production.Another measure of ecosystem maturity is the connectance index, whichexpresses the actual number of links in a food web to the total number of possiblelinks within the system (Table  5).  Odum (1969, 1971) expected systems to shiftfrom linear food chains to more complex web-like structures as they matured.Thus the 1950s model should have a higher connectance index than the 1980smodel, but the two measures are virtually identical (Table 4). The connectanceindex may be useful for comparing two time periods but is not useful whencomparing different ecosystems because it is affected by the number of boxes ina model.  In most systems, the actual number of links is roughly proportional tothe number of groups.  Moreover, the connectance index is strongly determinedby our level of diet details.A more useful comparative index of system complexity is the ‘system omnivoryindex’ (Christensen 1995).  This is the average omnivory index of all consumersweighted by the food intake.  It measures the distribution of feeding interactionsbetween trophic levels and can characterize the extent to which a systemdisplays web like features.  An individual group with an omnivory index of zero isa specialized feeder.  High values indicate the group feeds on organisms at manydifferent trophic levels.  The omnivory index was 0.183 in the 1950s and 0.157 inthe 1980s, suggesting that the eastern Bering Sea had more complex feedinginteractions in the 1950s (Table  5).   The lower value of this index in the 1980smodel compared to the 1950s appears to be due a difference in our assumptionsabout the species composition and diet composition of benthic epifauna, whosediet in the 1980s model was assumed to have a higher proportion of detrituscompared to the 1950s model. These omnivory estimates are consistent withthose calculated from other ecosystems (Table 9) such as the Alaska Gyre andBritish Columbia Shelf (0.103-0.140) and the Bolinao and Virgin Islands coralreefs (0.182-0.227) which have a complex web structure (Aliño et al. 1993, andOpitz 1993).34‘Total system throughput’ is another descriptive statistic of an ecosystem.  It is thesum of all flows from consumption, export, respiration and detritus, andrepresents the ‘size’ of the entire system in terms of its flow. Throughput wasgreater in the 1950s than in the1980s model (6,534 versus 5,691 t•km-2 year-1),again primarily due to our assumption of a much higher infauna biomass in the1950s than in the 1980s.  It is noteworthy that both of values for the Bering Seaare greater than other studied shelf systems except for those areas with strongupwelling (Table 9).  The fraction of the total system throughput that is recycledwithin the system was greater in the 1950s (13.2%) than in the 1980s (11.1%).With the exception of Brunei, this is more than three times greater than mostother shelf systems listed in Table 9.    The cycling index generally increases assystems mature (Christensen and Pauly 1998).  This is an indication that theBering Sea is more mature relative to other shelf ecosystems.Ascendancy measures the average mutual information in a system and is scaledby throughput (Christensen and Pauly 1992a,b). Taking the difference betweentotal system capacity and ascendancy is a measure of system overhead.Overhead provides limits on how much the ascendancy can increase and is areflection of the systems ‘strength in reserve’ from which it can draw to meetunexpected perturbations (Ulanowicz 1986).The eastern Bering Sea has a relatively low ascendancy when compared to othershelf systems. The system overhead is approximately 60-65% (with a capacity of100% and an ascendancy of approximately 30-35%).  This suggests that theBering Sea ecosystem has significant ‘strength in reserve’. Unfortunately it isunclear whether strength in reserve means resilience or resistance (i.e., thesystem may either be resistant to perturbations or it may be resilient and ‘bounceback’ quickly. It may even have a combination of both qualities).Ecosim Model ResultsWe used Ecosim (Walters et. al 1997) to investigate how fishing and a regimeshift might have affected the eastern Bering Sea ecosystem. Ecosim is anextension of Ecopath that can run dynamic and equilibrium simulations.Equilibrium simulations dynamically adjust the ecosystem to compensate forchanges in fishing mortality rates.  The simulations calculate the equilibriumbiomass for all species in the ecosystem over a range of fishing mortality that isdirected at one or more species.  Dynamic simulations can change fishingmortality rates, and can follow how the effect of fishing one group is propagatedover time through all the others. Dynamic simulations can also simulate a regimeshift by changing primary productivity and diet compositions.  Assumptions aboutthe available flow of food from one group to another can be evoked in bothdynamic and equilibrium models to simulate ‘top-down’ or ‘ bottom-up’ controlEcosim represents trophic ontogeny by linking adult and juvenile groups with adelay-differential model structure.  It simultaneously accounts for the numbersand biomass of fish in each group by setting the rates of flow (recruits) from thejuvenile to the adult boxes, and from the adult to the juvenile boxes (recruitment)35(Walters et al. 1997).  We applied the two-pool delay differential model structureto pollock to capture the effects of trophic ontogeny, cannibalism, and distinctroles that adult and juvenile pollock play in the eastern Bering Sea.  Model inputparameters were based on Bakkala  (1993) and FishBase 98 (Froese and Pauly1998), and included: age at graduation to the adult box (2 years), the meanweight at graduation (0.052 Kilos, 20 cm fish), and the von Bertalanffy growthparameter (k = 0.18 year-1) (although it should be noted that more recent dataspecific to the eastern Bering Sea indicate weight of age-2 pollock isapproximately 170g and the k parameter is around .228 (Wespestad et al. 1996)).We applied equilibrium simulations to 5 species groups: baleen whales, spermwhales, pelagic fishes, pollock and large flatfish.  The goal was to determine howthe biomass of the 25 groups of species making up the Bering Sea might respondto changes in fishing pressure applied to each of these 5 species.  We changedfishing pressure on baleen whales, sperm whales and pelagic fish in the 1950smodel, and on pollock and large flatfish in the 1980s model.We used dynamic simulations to gain insight into the mechanisms that mighthave changed the ecosystem from its 1950s state to its 1980s state.  We alsoused the model to project the future consequences of changing some of thecurrent fishing policies.  The specific dynamic simulations we ran included:· A regime shift simulation that altered the amount of primary production in theBering Sea and shifted the diet of many predators from pelagic fishes topollock.  The simulation was initiated with the 1950s Ecopath model and runfor 30 years. Results were compared with the observed status of the systemas detailed in the 1980s model;· A fishery simulation that explored whether human harvesting alone couldexplain the changes that occurred over 30 years;· Three pollock fishing scenarios that explored how the ecosystem mightchange from its 1980s state if pollock fishing was changed.  We consideredwhat might happen to the Bering Sea if pollock fishing decreased thebiomass of adult pollock by 50%.  We also considered what might happen ifpollock were overfished to the point that all the juvenile and adult pollockwere removed.  Finally, we considered how the Bering Sea ecosystem mightlook if pollock were not caught at all.  All simulations were run for 30 years;· Finally, we tried altering fishing mortality of pelagics and pollock (mid-trophiclevel groups) and estimating the amount of time required for the system torecovery.  Fishing mortality (F) on these two groups was drastically increased(20 times fold) for a period of 10 years, then released back to the original F(Table 10).  Simulations were run for 100 years using both the 1950s and1980s Ecopath models for initiation.36Table 10. Original and increased fishing mortality rates (year-1)used to assess system recovery time.Fished group1950 1980 1950 1980Pelagics 0.0030 0.0150 0.0600 0.3000Adult Pollock 0.0025 0.0700 0.0500 1.40001 x F 20 x FEquilibrium Simulation ResultsThere are five sets of results for the equilibrium simulations, which we present asa series of graphs.  The top left panels in each of the figures show the rate ofexploitation (F) used in the 1950s or 1980s Ecopath models (marked by anarrow) and how altering F affects the numbers present (biomass) and thebiomass caught (catch).  F is the instantaneous fishing mortality rate (Ricker1975) and is approximately equal to the fraction of the population that is removedeach year over the range of values we considered (0.00 – 0.30 year-1).The exploitation rate of baleen whales in the 1950s model was about 12% peryear (Fig. 9).  Increasing F resulted in a short-term increase in numbers caught,but an ultimate decline in harvest as the population decreased to zero.  ReducingF increased the whale population and ultimately resulted in a greater sustainableharvest.Removing baleen whales from the 1950s Bering Sea model increases thetoothed whales, sperm whales, walruses, bearded seals, sea lions and sea birds(Fig. 9). Reducing baleen whales increased zooplankton biomass (reducedpredation) and increased their major competitors (pollock and cephalopods)which are fed upon by other marine mammals.  Removing baleen whales had apositive effect on pollock but no discernable effect on pelagic fishes or seals(northern fur seals, harbor seals, larga seals, ribbon seals and ringed seals),which were the next most important competitors of baleen whales.The model predicts that increases of baleen whales in the eastern Bering Seacould significantly reduce the abundance of pollock, cephalopods and deepwaterfishes through direct competition for zooplankton.37          Baleen Whales – 1950sa. b.c. d.e. f.-20-1001020% change biomassToothedSpermWal  &BeardSealsSteller0246810120.00 0.05 0 .10 0.15 0.20 0.25 0.30F-B aleen whalesBiomass (t/km-2) (t/km-2)BiomassCatch-50-40-30-20-1001020304050% change biomassPisc. BirdsAd. PollockJuv. PollockPelagicsCephalopods-20-1001020% change biomassO.Dem fishL. flatfishSm. FlatfishDeepwater fish-10010% change biomassL.zoopsH.zoopsPhytesJellyfish-20-1001020%change biomassB.part.feedInfaunaEpifaunaFig. 9.  Equilibrium biomass for Bering Sea species following changes to the fishingmortality of baleen whales in the 1950s model.  Arrows mark the instantaneous rateof fishing (F; year-1) during the 1950s.  The top left panel shows changes in thebiomass and catch of baleen whales under different levels of F.  The other 5 panelsshow the relative change (%) that could occur to other species in the ecosystem tocompensate for changes in the abundance of baleen whales.38Reducing sperm whale abundance in the 1950s model doubles the deepwater fishpopulation, and increases the cephalopods (Fig. 10). These are the two major prey of spermwhales.  Increasing sperm whale prey reduces Steller sea lions, sea birds, flatfish and pelagicfishes.  Sperm whales presumably exert some control on cephalopods and deep waterfishes, which affect other species in the food web.Sperm  Whales – 1950sa. b.c. d.e. f. 0 .05 0 .10 0.15 0.20 0 .25F-Sperm whalesBiomass (t/km-2) (t/km-2)BiomassCatch-20-1001020% change biomassB aleenToothedW al &BeardS ealsS teller-30-20-100102030% change biomassP isc. BirdsA d. PollockJuv. PollockP elagicsCephalopods-120-100-80-60-40-20020406080100120% change biomassO.Dem fishL. flatfishSm. FlatfishDeepwater fish-10010% change biomassL.zoopsH.zoopsP hytesJellyfish-10010% biomass changeB .part.feedInfaunaE pifaunaFig. 10.  Equilibrium biomass for Bering Sea species following changes to thefishing mortality of sperm whales in the 1950s model.  Arrows mark theinstantaneous rate of fishing (F; year-1) during the 1950s.  The top left panel showschanges in the biomass and catch of sperm whales under different levels of F.  Theother five panels show the relative change (%) that could occur to other species inthe ecosystem to compensate for changes in the abundance of sperm whales.39Pelagic fish play a central role in the diet of many groups in the 1950s model andcan have large system wide effects (Fig. 11).  Reducing pelagic fish in this modelcaused significant declines of piscivorous birds, Steller sea lions and large flatfishthat fed on them.  The model predicts an 80% decline in sea lion abundance witha 50% decline in pelagic abundance.  Food groups (zooplankton) andcompetitors (pollock) of pelagic fishes increased when they were released.Some trophic cascade effects were observed, such as the increase in benthicparticulate feeders due to an increase in food (zooplankton) and a decrease inpredators (large flatfishes, walruses and bearded seals).  Pelagic Fish – 1950sa. b.c. d.e. f.0510152025300.00 0 .10 0 .20 0.30 0.40 0.50F-PelagicsBiomass (t/km-2)0123456Catch (t/km-2)BiomassCatch-100-80-60-40-20020406080100% change biomassBaleenToothedSpermW a l &BeardSealsSteller-100-80-60-40-20020406080100% change biomassP isc. BirdsA d. PollockJuv. PollockCephalopods-100-80-60-40-20020406080100% change biomassO.Dem fishL. flatfishSm. FlatfishDeepwater fish-40-2002040% change biomassL.zoopsH.zoopsPhytesJellyfish-80-60-40-20020406080% change biomassB .part.feedInfaunaE pifaunaFig. 11.  Equilibrium biomass for Bering Sea species following changes to thefishing mortality of pelagic fishes in the 1950s model. Arrows mark theinstantaneous rate of fishing (F; year-1) during the 1950s.  See Fig. 9 for furtherexplanations.40Increasing fishing pressure on pollock has a very small, or minimal effect on theadult biomass in the 1980s model.  This is apparently due to continuousreplenishment from the juvenile population (which increases due to lowerpredation pressure – i.e., cannibalism). System wide effects are therefore minimalbecause the adult population does not change appreciably.  The model predictsthat seals, sea lions and piscivorous birds would increase if more adult pollockwere caught because the abundance of juvenile pollock increases as cannibalismby adult pollock is reduced.         Pollock – 1980sa. b.c. d.e. f.051015202530350 0.1 0.2 0.3 0.4 0.5F-PollockBiomass (t/km-2)02468101214161820Catch (t/km-2)BiomassCatch-10010% change biomassBaleenToothedSpermW al &BeardSealsSteller-20-1001020% change biomassPisc. BirdsJuv. PollockPelagicsCephalopods-20-1001020% change biomassO.Dem fishL. flatfishSm. FlatfishDeepwater fish-20-1001020% change biomassL.zoopsH.zoopsPhytesJellyfish-30-20-100102030% change biomassB.part.feedInfaunaEpifaunaFig. 12.  Equilibrium biomass for Bering Sea species following changes to thefishing mortality of adult pollock in the 1980s model. Arrows mark theinstantaneous rate of fishing (F; year-1) during the 1980s.  See Fig. 9 for furtherexplanations.41Fishing large flatfish in the 1980s model increases Steller sea lions, seabirds anddeepwater fish due to a competitive release of their food (pollock, small flatfishand other demersal fish).  The two groups of species that do not benefit fromincreased fishing pressure on large flatfish are seals and crabs (Fig. 13).Cascading effects resulting from a decrease in large flatfishes include a decline insome pinnipeds (predators of flatfish), which further reduces predation ondeepwater fish.Large Flatfish –1980sa. b.c. d.e. f. 0.10 0.20 0.30 0.40F-Large flatfishBiomass (t/km-2) (t/km-2)BiomassCatch-20-1001020% change biomassBaleenToothedSpermW al &BeardSealsSteller-50-40-30-20-1001020304050% change biomassPisc. BirdsAd. PollockJuv. PollockPelagicsCephalopods-40-30-20-10010203040% change biomassO.Dem fishSm. FlatfishDeepwater fish-20-1001020% change biomassB.part.feedInfaunaEpifauna-10010% change biomassL.zoopsH.zoopsPhytesJellyfishFig. 13.  Equilibrium biomass for Bering Sea species following changes to thefishing mortality of large flatfish in the 1980s model. Arrows mark theinstantaneous rate of fishing (F; year-1) during the 1980s.  See Fig. 9 for furtherexplanations.42Dynamic Simulation ResultsWe ran five sets of dynamic simulations.  The first began in 1955 and simulatedthe removal of reported catches of fish over a 30-year period. We calculatedfishing mortality as F = C/B (Table 11). With the exception of baleen and toothedwhales, we assumed a simple linear change in the fishing mortality between the1950s and 1980s (a 30-year period). For baleen and toothed whales, fishingmortality was held constant for the first 10 years, after which it declined linearly tozero by 1970.Table 11.  Estimated biomass, harvest and fishing mortality in the 1950s and 1980s.Biomass Harvest Fishing Biomass Harvest Fishingt•km-2•y-1 t•km-2•y-1 Mortality t•km-2•y-1 t•km-2•y-1 MortalityB C F=C/B B C F=C/BBaleen whales 0.696 0.084 0.121 0.394 0.000 0.000Toothed whales 0.009 0.000 0.000 0.009 0.000 0.000Sperm whale 0.439 0.021 0.048 0.208 0.000 0.000Beaked whales 0.001 0.000 0.000 0.001 0.000 0.000Walrus & bearded 0.054 0.006 0.111 0.074 0.004 0.054Seals 0.106 0.005 0.047 0.066 0.001 0.015Steller sea lion 0.029 0.001 0.034 0.019 0.001 0.053Pisc. Birds 0.006 0.000 0.000 0.006 0.000 0.000Adult pollock 5.500 0.014 0.003 27.451 1.895 0.069Juv. Pollock 0.942 0.000 0.000 6.000 0.000 0.000O. demer. Fish 8.957 0.001 0.000 3.904 0.128 0.033Large flatfish 1.169 0.002 0.002 1.900 0.050 0.026Small flatfish 8.530 0.105 0.012 9.161 0.211 0.023Pelagics 28.869 0.083 0.003 13.644 0.212 0.016Deepwater fish 1.011 0.001 0.001 0.407 0.007 0.017Jellyfish 0.048 0.000 0.000 0.048 0.000 0.000Cephalopods 3.500 0.000 0.000 3.500 0.000 0.000Benth.P. feeders 29.000 0.010 0.000 5.800 0.108 0.019Infauna 75.000 0.000 0.000 46.500 0.000 0.000Epifauna 8.000 0.000 0.000 5.858 0.000 0.000Large Zoops 44.000 0.000 0.000 44.000 0.000 0.000Herb. Zoops 55.000 0.000 0.000 55.000 0.000 0.000Phytoplankton 32.000 0.000 0.000 32.000 0.000 0.000Discards 0.000 0.000 - 0.000 0.000 -Detritus - - - 0.000 0.000 -1950s 1980s43Fishing (i.e., killing whales and catching fish) had little effect on the simulatedecosystem and failed to produce the large abundance of pollock observed in the1980s.  The only way to dramatically increase the amount of pollock and largeflatfish in our simulated ecosystem was to abruptly change the amount of primaryproduction in the Bering Sea.       Commercial Fishing Simulationa. b.c. d.e.-200-150-100-500501001502000 10 20 30% change biomassBaleenToothedSpermWalrus+beardedSealsSteller-25-20-15-10-505101520250 10 20 30% change biomassBirdsAd.PollockJ.PollockPelagicsCephalopods-50-40-30-20-10010203040500 10 20 30% change biomassO.Dem.FishL.flatfishSm.flatfishDeepwater fish-10-8-6-4-202468100 10 20 30% change biomassBenth.part.feedersInfaunaEpifauna-10-8-6-4-202468100 10 20 30% change biomassJellyfishL.zoopsHerb.zoopsPhtyesFigure 14.  Dynamic simulation of the effects of commercial fishing from the1950sto the 1980s.  Note that some predators switched from eating pelagics to eatingpollock, and that fishing mortality for Baleen whales was increased three-foldbased on the EE values being roughly 3 times too high to balance the systemduring the 1950s.44Venrick et al. 1987 (see Fig. 3.16 in NRC 1996) estimate that the amount ofchlorophyll (an index of phytoplankton production) in the central North Pacificalmost doubled between the 1960s and the 1980s.  The timing of this change inthe 1970s is consistent with an hypothesized ‘regime shift’.   We thereforesimulated an increase in primary productivity based on reported changes in theindex of phytoplankton production detailed in Figure 3.16 of the NRC report(1996).Adding more primary production to the 1950s ecosystem model increased theamounts of large zooplankton and cephalopods available as prey to other species(Fig. 15).  Increases in their numbers led in turn to sharp cyclical increases anddecreases in the abundance of fish and invertebrates, which dampened overtime.  Groups such as marine mammals do not have the capacity to rapidly turnthe increased production into biomass and were not strongly affected by thesharp cycles in prey abundance.  Curiously, only the Steller sea lions did notshow an overall increase in numbers when primary production was increased.The effects of simulating an increase in primary production were propagatedthrough the food web. Pelagic fishes, benthic particulate feeders, sea birds andmost mammals responded favorably to increased primary production (Fig. 15,Table 12).  However, this simulation failed to completely move the system from its1950s state to its 1980s state.A change in primary production can explain more of the changes that occurred inthe ecosystem than can the effects of commercial fishing.  However, the regimeshift would have had to affect more than just primary production to move thesimulated ecosystem from our hypothesized 1950s state to our more certainunderstanding of the Bering Sea in the 1980s.  Either something had to happen inconcert with an increase in primary production to favor the survival of pollock andlarge flatfish, and disfavor the pelagics and benthic particulate feeders, or we mayneed to re-examine our evidence regarding the state of the Bering Sea in the1950s, particularly with regard to pollock abundance and trophic connections topollock.Pollock appear to affect the abundance of crabs, cephalopods, pelagic fishes andother demersal fishes.  In the absence of pollock (Fig. 16), these speciesincreased as did marine mammals.  The extinction of pollock could lead to thecollapse of large flatfish in the model.  However, this is contrary to evidence fromstock assessments that showed Greenland turbot at its highest abundance in the1970s prior to the large increase in pollock populations seen in the 1980s.  Ourmodels suggests the abundance of large flatfish is tied to the abundance ofpollock.  Similarly, piscivorous sea birds are dependent on juvenile pollock andwould also face sharp declines if they did not switch to another major prey (Fig.16).It has been argued that the Steller sea lion population would increase if pollockfishing were curtailed. Presumably, there would be more pollock for Steller sealions to eat.  We simulated forward from the 1980s and explored this possibility bystopping all fishing of pollock (Fig. 17).  Surprisingly, Steller sea lions and othermarine mammals did not respond favorably to increases in the abundance of45adult pollock.  Increases in adult pollock resulted in decreases in the abundanceof juvenile pollock that piscivorous birds and many mammals fed upon.Thus, paradoxically, the model suggests that one means to increase Steller sealion populations may be to fish pollock harder.  A positive increase in sea lionnumbers was seen when pollock fishing was increased to cause a 50% reductionin adult pollock biomass (Fig. 18).  Reducing adult pollock increased theabundance of juvenile pollock by reducing cannibalism.  This in turn increased theabundance of crabs, birds, pelagic fish, other demersal fish and marinemammals.    1950s – Fishing & Regime Shifta. b.c. d.e.-160-120-80-40040801201600 10 20 30% change biomassBaleenToothedSpermWalrus+beardedSea lsSteller-300-250-200-150-100-500501001502002503000 10 20 30% change biomassBirdsAd.PollockJ.PollockPelagicsCephalopods-150-100-500501001500 10 20 30% change biomassO.Dem.FishL.flatfishSm.flatfishDeepwater f ish -200-150-100-500501001502000 10 20 30% change biomassBenth.part.feedersInfaunaEpifauna-360-300-240-180-120-600601201802403003600 10 20 30% change biomassJellyfishL.zoopsHerb.zoopsPhtyesFigure 15. A dynamic simulation that began in the 1950s and incorporated realisticcommercial catches (of fish and whales) and a regime shift (a cyclic increase inprimary production).  Diets of some predators were allowed to switch from pelagicsto pollock, and the harvesting of baleen whales was increased by a factor of three.46Table 12.  Results from dynamic simulations showing how historic fishing patternswould have changed the biomass of 25 groups of species between the1950s modelstate and the 1980s, and how the biomass would have theoretically changed if aregime shift had increased primary production alone.  The table shows the percentchange observed in biomass between the 1950s and the 1980s, and the percentchange observed when the model included historic fishing and the effects of aregime shift.  Direction of change is indicated by + for increase, - for decrease and 0for no appreciable change.Functionalgroup Observed ObservedFishing Shift Fishing ShiftBaleen whales - + - -43 7 -71Toothed whales 0 0 0 0 0 6Sperm whales - + + -53 189 179Beaked whales 0 0 0 0 -2 8Walrus & bearded + + + 37 133 158Seals - + + -38 65 87Steller sea lion - - 0 -34 -11 0Pisc. Birds 0 + + 0 19 98Adult pollock + 0 + 399 -3 23Juv. pollock + 0 + 537 1 18O. demer. fish - - + -56 -12 16Large flatfish + 0 + 63 -2 67Small flatfish + + - 7 18 -16Pelagics - + + -53 14 120Deepwater fish - - - -60 -56 -16Jellyfish 0 0 + 0 2 196Cephalopods 0 - - 0 -20 -55Benth.P. feeders - - - -80 -8 -19Infauna - 0 + -38 2 21Epifauna - 0 0 -27 4 6Large zoops 0 0 0 0 -3 7Herb. zoops 0 0 + 0 3 74Phytoplankton 0 0 + 0 -1 39Simulated SimulatedDirection of Change Change (%)Other retrospective analyses have also incorporated cannibalism into agestructured models of pollock.  They indicate that although cannibalism issignificant and can explain some of the observed variation in juvenile survivalrates, there is a large unexplained component of recruitment variation that is likelylinked to environmental factors (Livingston and Methot 1998; Livingston andJurado-Molina, in press).  These environmental factors have the potential to shiftadult pollock abundance to very high levels, as was observed with the 1978pollock year class.  The results of forward projections of the effect of fishing on theecosystem using ECOSIM or age-structured models such as MSFOR, relyheavily on our assumptions about recruitment (Gislason 1993).  Sensitivity of theresults here need to be examined in that light.  A long-term research goal shouldbe to incorporate the results of process-oriented research on climatic links torecruitment variation into models which presently focus only on the predation linksbetween species.47     1980s – 100% Decrease of Adult Pollocka. b.c. d.e.-30-20-1001020300 10 20 30% Change biomassBaleenToothedSpermW&BSealsSteller-100-80-60-40-200204060801000 10 20 30% Change biomassBirdsPollock APelagicsCephalop-100-80-60-40-200204060801000 10 20 30% Change biomassO.D.FishL. FlatfishS. FlatfishDeep. Fish-30-20-1001020300 10 20 30% Change biomassL. ZoopH. ZoopPhyto-100-80-60-40-200204060801001200 10 20 30% Change biomassBenth. FeedInfaunaEpifaunaFig. 16.  A dynamic simulation showing the effect of removing adult pollock fromthe 1980s Bering Sea model ecosystem.  The model predicts the decline of pollockwould also lead to the decline of seabirds because they can no longer feed onjuvenile pollock.  Whether this would in fact happen depends upon whetherseabirds would switch their diet to the growing pelagic fish populations.  Therelease of food (zooplankton) allows pelagics and benthic particulate feeders toincrease, as well as their predators who benefit from reduced predation from acommon predator (large flatfish).48       1980s – 0% Decrease of Adult Pollocka. b.c. d.e.-100100 10 20 30% change biomassBaleenToothedSpermW&BSealsSteller -100100 10 20 30% change biomassBirdsPollock APollock JPelagicsCephalopd-100100 10 20 30% change biomassO. D. FishFlatfish LFlatfish SDeep. Fish-100100 10 20 30% change biomassBenth. FeedInfaunaEpifauna-100100 10 20 30% change biomassJellyfishZoop. LZoop. HPhytoFig. 17.  A dynamic simulation showing the effect of stopping commercial fishingfor adult pollock in the 1980s Bering Sea model ecosystem.  Catching no adultpollock results in a larger adult population and a smaller juvenile pollockpopulation.  The majority of other groups in the Bering Sea are largely unaffectedby a reduction in pollock fishing.49     1980s – 50% Decrease of Adult Pollocka. b.c. d.e.-20-10010200 10 20 30% change biomassBaleenToothedSpermW&BSealsSteller-70-60-50-40-30-20-100102030405060700 10 20 30% change biomassBirdsPollock APollock JPelagicsCephalopdSteller-40-30-20-100102030400 10 20 30% change biomassO. D. FishFlatfish LFlatfish SDeep. Fish-40-30-20-100102030400 10 20 30% change biomassBenth. FeedInfaunaEpifauna-30-20-1001020300 10 20 30% change biomassJellyfishZoop. LZoop. HPhytoFig. 18.  A dynamic simulation showing the effect on the 1980s Bering Sea modelecosystem of decreasing the biomass of adult pollock by 50%.  The model predictsan increase in the prey of adult pollock (large zooplankton and juvenile pollock), anincrease in competitors (mainly the fast growing groups – jelly fish, benthicfeeders, pelagics and sea birds).  Marine mammals show a positive change, but arelimited by their inherent low productivity.50System Recovery TimeIncreasing the fishing mortality by 20-fold on pelagics and pollock for 10 yearsbefore releasing it back to its original level affected all trophic levels in both the1950s and 1980s models.  Species that took the longest to return to their originalabundance were those at high trophic levels (Table 13).Fishing pollock heavily in the 1950s would not have affected many speciesbecause we assumed that overall pollock abundance was low and that nospecies were eating substantial amounts of pollock at that time.   Pelagic fishwere important in both time periods but seem to be more critical in the 1980s thanin the 1950s given that more species were impacted for longer periods by thelower pelagic fish abundance in the 1980s.  This may reflect the relatively lowbiomass of pelagics in the 1980s and the greater amount of time required forsmall populations of marine mammals to rebuild.Table 13. The last species to recover from an increase in fishing pressure onpelagics or pollock.  Fishing mortality (F year-1) on the two mid-trophic level groupswas drastically increased (20 times fold) for a period of 10 years, then releasedback to the original F (see Table 10).  We ran four simulations for 100 years initiatedwith the 1950s and 1980s Ecopath models. The table shows the number of years ittook for the slowest species to recover and the percentage of recovery achieved inthat time, or the percentage of recovery following 100 years of simulation.Model Grouped Species Fished Affected years %1950s Pelagics Steller sea lion 100 981950s Adult Pollock Adult pollock 23 1001980s Pelagics Steller sea lion 100 95Deepwater fish 100 97Seals 100 101Sperm Whales 100 1011980s Adult Pollock Steller sea lion 100 101Deepwater fish 100 98Seals 100 101Sperm whales 100 100Beaked whales 100 99Recovery51Discussion of Simulation ResultsEcosim tracked trophic interactions over 30-100 years of simulation.  It showedhow altering the abundance of one species can affect others, and how the systemas a whole might respond.  It is therefore a useful tool for understanding what rolecommercial fisheries may play in restructuring the Bering Sea.All of our simulations showed that, if our assumptions about the state of theBering Sea in the 1950s are accurate, trophic interactions alone cannot explainthe magnitude of changes that occurred in the biomass of major groups in theeastern Bering Sea between the 1950s and the 1980s.   This conclusion aboutthe Bering Sea differs from that drawn for the Gulf of Thailand, where fishing ratesalone could move the system from one state to another (1960s-1980s: before andafter the development of trawl fisheries, Christensen 1998).  These findings begthe question of whether tropical marine ecosystems are more influenced bytrophic interactions than environmental events compared to northern marineecosystems.Our models suggest that removing historic levels of commercial fish catches fromthe Bering Sea had little affect on the dynamics of the system.   A regime shiftthat affected the base of the food web (primary production) would have had apronounced effect on the abundance of many species, but is also insufficient byitself to explain the totality of changes that occurred.  Some other factor wouldhave had to be at play to favor the survival of certain species (such as pollock)over others (such as crabs).  That factor may be physical oceanographic changesin water temperature and ocean currents that increase survival rates of certainspecies such as pollock and other groundfish.Strong year classes of groundfish were more frequent in the eastern Bering Seaand Gulf of Alaska after 1976, and are linked to large scale physical forcingfactors (Hollowed et al. 1998).  Similarly, Quinn and Niebauer (1995) linkedpollock survival to both environmental and biological factors such as predation.An age-structured model of walleye pollock that includes predation also providesevidence of higher juvenile pollock survival in the years after 1978, in early lifehistory stages before predation mortality occurs (Livingston and Methot 1998).Thus, there is evidence of climate-related shifts in survival rates of pollockbeginning in the late 1970s.Another factor influencing our inability to project forward from the 1950s model toattain a semblance of the 1980s Bering Sea ecosystem, is our uncertainty overthe state of the Bering Sea in the 1950s.  We lacked quantitative data on manyspecies in the 1950s and assumed that they had the same abundance as wasobserved in the 1980s.  For other species, we used data from Soviet bottom trawlsurveys in the 1950s that suggested low pollock abundance relative to thepresent.  However, there are large uncertainties in comparing those historicalcatch rates to present day catch rates due to the different gear used by theSoviets in the early days compared to the gear presently used in U.S surveys.Anecdotal evidence in reports of these Soviet fisheries investigations indicatesthat our assumptions about low pollock biomass and little predation on pollock byother species may be inaccurate.  Shuntov (1972) reported that during their52investigations from 1957-1964, walleye pollock was one of the most commonBering Sea fishes and was a staple food of large flatfish as well as other fishes.However, stomach contents of Steller sea lions shot in the Gulf of Alaska show ashift in diet from largely pelagic fishes in the 1950s to pollock and flatfish in the1970s and 1980s (Alverson 1992).  Future work to improve the 1950s modelshould include testing alternative assumptions about the dominance of pollockduring that time period.ConclusionsOne of the questions we posed at the outset of this study was whether the nicheonce occupied by whales is now occupied by pollock.  We took two approaches.One was to remove pollock from the 1980s ecosystem to see if whalepopulations would rebuild.  The other was to remove the whales from the 1950secosystem to see if pollock would explode.  Neither approach producedconclusive results about the role that whaling played in increasing pollockbiomass.  The most we can say is that removing whales would have had apositive effect on pollock by reducing competition for food.  However, whalingalone is insufficient to explain the 400% increase in adult pollock biomass thatmay have occurred between the 1950s and the 1980s.Although we did not focus on the relative importance of pelagics and pollock inthe eastern Bering Sea ecosystem, there appears to have been a switch betweendominance of these two mid-trophic level species.  This switch between the1950s and 1980s merits further investigation.The magnitude of changes in the biomass of the major groups in the easternBering Sea cannot be explained solely through trophic interactions. The key toexplaining the large-scale dynamics of the eastern Bering Sea may well be anenvironmental shift that favors one complex of species over another.  Our modelssuggest that oceanographic factors such as changes in water temperature orocean currents must have been at play and may be an important factor inaffecting ecosystem production and recruitment variation in species.  It isintriguing to think that the Bering Sea may have two alternative states, containingtwo suites of dominant species, and that the transition between the two may bevery rapid.  Steele and Henderson (1984) and Spencer and Collie (1996)describe models that show the possibility of such rapid transitions.One of the more interesting predictions of the ecosystem model is that stoppingpollock fishing might negatively affect many of the top predators in the BeringSea.  The model indicates that top predators might realize a greater benefit ifpollock were fished at high levels or if large flatfish such as arrowtooth flounderwere fished at higher rates due to the competitive release of prey.  We are notadvocating such changes in fishing, but rather are pointing out that the system,when viewed at this aggregated level, may not respond in the way that many53people might have predicted without the assistance of a quantitative, holisticmodel.We recognize that our assumptions about recruitment influence our conclusionsabout the effects of fishing on the system, and that species likely responddifferently to climate forcing at the inter-annual and inter-decadal time scales.Although we attempted to capture recruitment variability due to predation, theclimate related variability is large and could show that the system may notrespond in a way that our predation model indicates. Thus, future ecosystemsimulations should explicitly consider the sensitivity of model results to differentassumptions about climate effects on recruitment.Our models do not yet capture the spatial aspects of foraging and fishingremovals that may be important in explaining marine mammal and fisheryinteractions (Trites et al., in press).  More research is needed to characterize theforaging ecology of Steller sea lions and the seasonal changes in the distributionand abundance of pollock and sea lions relative to fishing, to fully evaluate theeffect of fishing on sea lions.  Further scrutiny should also be given to ourassumption of cannibalism because of the fundamental role it appears to play instructuring the Bering Sea ecosystem.While some indices indicated that the 1950s ecosystem, as we modeled it, was ata more mature state than in the 1980s, it cannot be said that the 1980s systemwas unhealthy.  Rather, the indices pointed out the large role that benthic infaunaand epifauna play in determining ecosystem-level estimates of maturity andhighlighted our uncertainty about these parts of the eastern Bering Seaecosystem.   There is still a great deal of uncertainty about the state of the BeringSea in the 1950s that may require further analysis and model experiments.Our models of the eastern Bering Sea ecosystem are an important first step indeveloping an ecosystem framework that will guide fisheries managers.  Theobvious next steps are to complete validation of these models and to develop aspatially explicit model that can track the movements of species from one area ofthe Bering Sea to another.  Another critical need is to incorporate habitat andenvironmental data, or perhaps to link Ecopath and Ecosim to climate models(Trites et al., in press).  A fuller understanding of the eastern Bering Sea is onlylikely to come when biological models are ultimately linked to physical models.54Literature CitedAlaska Department of Fish and Game.  1994.  Annual management report forthe shellfish fisheries of the westward region, 1992.  Regional InformationReport No. 4K94-9.  Alaska Department of Fish and Game, CommercialFisheries Management and Development Division.  211 Mission Road,Kodiak, AK 99615.Aliño, P.M., L.T. McManus, J.W. McManus, C.L. Nañola, Jr., M.D. Forbes, G.C.Tronno, Jr. And G.S. Jacinto. 1993. Initial parameters estimations of acoral reef flat ecosystem in Bolinao, Pangasinan, Northwestern Phillipines,p. 252-258. In V. Christensen and D. Pauly (eds.) Trophic models ofaquatic ecosystems. ICLARM Conf. Proc. 26, 390 p.Allen, K.R. 1971.  Relation between production and biomass.  J. Fish. Res. Bd.Canada. 28:1573-1581.Alverson, D.L. 1992. Commercial fisheries and the Steller sea lion(Eumetopias jubatus): the conflict arena.  Rev. Aquat. Sci. 6:203-256.Alverson, D. L. and Hughes, S.E. 1995. Bycatch: from emotion to effectivenatural resource management. p. 13-29. In Solving Bycatch:Considerations for today and tomorrow. Alaska Sea Grant CollegeProgram Report No. 96-03, University of Alaska Fairbanks.Arai, M.N. 1997.  A Functional Biology of Scyphozoa.  Chapman & Hall, London.316 p.Baker, C.S. and L.M. Herman 1987. Alternative population estimates ofhumpback whales (Megaptera novaeangliae) in Hawaiian waters. Can J.Zool. 65: 2818-2821.Bakkala, R.G. 1993. Structure and historical changes in the groundfish complexof the eastern Bering Sea. U.S. Department of Commerce, NOAA Tech.Rep. NMFS 114. 91 pp.Banse, K. and S. Mosher. 1980.  Adult body mass and annualproduction/biomass relationships of field populations.  Ecol. Monogr.50:355-379.Barnes, R.D. 1980.  Invertebrate Zoology. Saunders College, Philadelphia. 1089p.Barrett-Lennard, L.G., K. Heise, E.Saulitis, G. Ellis, C. Matkin.  1995. The impactof killer whale predation on Steller sea lion populations in British Columbiaand Alaska. University of British Columbia, Fisheries Centre, 2204 MainMall, Vancouver, B.C. V6T 1Z4.55Berzin, A.A. 1959.  On the feeding of sperm whales (Phy eter catodon) in theBering Sea. [in Russian] Izv. Tikhookean. Nauchno-Issled. Inst. Rybn.Khoz. Okeanogr. (TINRO) 47: 161-165. [translation available from NationalMarine Mammal Laboratory, NMFS, Seattle WA).Berzin, A.A., and A.A. Rovnin. 1966. The distribution and migrations of whales inthe notheastern part of the Pacific, Chukchee and Bering Seas. Pages103-136, In. K.I. Panin (ed.), Soviet research on marine mammals of theFar East. U.S. Dep. Interior Bureau of Commercial Fisheries, Seattle, WA.Bigg, M.A. 1990. Migration of northern fur seal (Callorhinus ursinus), on St. PaulIsland, Alaska. Fish Bull. 84: 383-394.Bonfil, R., G. Munro, U.R. Sumaila, H. Valtysson, M. Wright, T. Pitcher, D.Preikshot, N. Haggan, and D. Pauly (eds) 1998.  Distant water fleets: anecological, economic and social assessment. Fisheries Centre ResearchReports 6(6), 111pp.Braham, H.W. 1991.  Endangered whales: status update. A Report on the 5-yearstatus of stocks review under the 1978 amendments to the U.S.Endangered Species Act. NMFS Unpublished Report.Braham, H.W. 1984. The bowhead whale, Balaena mysticetus. Mar. Fish. rev.46: 45-53.Braham, H.W., J.J. Burns, G.A. Fedoseev, and B.D. Krogman. 1984. Habitatpartitioning by ice-associated pinnipeds: Distribution and density of sealsand walruses in the Bering Sea, April 1976. In F.H. Fay and G.A.Fedoseev (eds.). Soviet-American cooperative research on marinemammals. Vol. 1: Pinnipeds, p. 25-47. U.S. Dep. Commer., NOAA Tech.Rep. NMFS-12.Brethes, J-C., B. Parent, and J. Pellerin.  1994.  Enzymatic activity as an index oftrophic resource utilization by the snow crab Chionoecetes opilio (O.Fabricius).  J. Crust. Biol. 14:220-225.Brodeur, R.D. and P.A. Livingston, 1988.  Food habits and diet overlap of variouseastern Bering Sea fishes. NOAA Tech. Memo. NMFS F/NWC-127, 76p.Bukhtiyarov, Y.A., Frost, K.J. and Lowry, L.F. 1984. New information on foods ofthe spotted seal, Phoca largha, in the Bering Sea in spring. p. 55-59. InF.H. Fay and G.A. Fedoseev (eds.). Soviet-American cooperative researchon marine mammals. Vol. 1: Pinnipeds, U.S. Dep. Commer., NOAATechnical Report NMFS 12.Burns, J.J. 1973. Marine mammal report. Alaska Dept. Fish and Game, Pittman-Robertson Project Rep. W-17-3, W-17-4, and W-17-5.Burns, J.J. 1981a. Bearded seal - Erignathus barbatus Erxleben, 1777. p. 145-170, In S.H. Ridgway and R.J. Harrison (eds.), handbook of MarineMammals. vol. 2, Seals. Academic Press, New York.56Burns, J.J. 1981b. Ribbon seal - Phoca fasciata. p. 89-109. In S.H. Ridgway andR.J. Harrison (eds.), Handbook of Marine Mammals. vol. 2. Seals.Academic Press, New York.Burns, J.J. 1986. Ice seals. p. 216-229.  In D. Haley (ed.), Marine mammals ofeastern North Pacific and arctic waters, 2nd ed., rev.  Pacific SearchPress, Seattle, WA.Calambokidis, J., G.H. Steiger, J.M. Straley, T. Quinn, J. Barlow, L.M. Herman,S. Cerchio, D.R. Salden, M. Yamaguchhi, F. Sato, J.R. Urban, J.Jacobson, O.Von Zeigesar, K.C. Balcomb, C.M Gabriele, M.E. Dahlheim,N. Higashi, J.K.B. Ford, Y. Miayfumi, P. Ladron de Guevara, S.A. Mizroch,L. Schlender, and K. Rasmussen. 1997. Population abundance andstructure of humpback whales in the North Pacific basin. Draft FinalContract Report 50ABNF500223 to Southwest Fisheries Science Center,P.O. Box 271, La Jolla, CA 92038. 67 p.Calkins, D.G. and E. Goodwin. 1988.  Investigation of the declining sea lionpopulation in the Gulf of Alaska. Unpubl. Rep., Alaska Dep. Fish andGame, 333 Raspberry Road, Anchorage, AK 99518. 76 p.Christensen, V. 1995. Ecosystem maturity - towards quantification. Ecol.Modelling 77:3-32.Christensen, V. 1998. Fishery induced changes in a marine ecosystem: insightfrom models of the Gulf of Thailand.  Journal of Fish Biology 53 (suppl. A):128-142.Christensen, V. and D. Pauly. 1992a. ECOPATH II - a software for balancingsteady-state models and calculating network characteristics. Ecol.Modelling 61:169-185.Christensen, V. and D. Pauly. 1992b. A guide to the ECOPATH II program(version 2.1). ICLARM Software 6, 72 p.Christensen, V. and D. Pauly (Editors). 1993. Trophic models of aquaticecosystems. ICLARM Conference Proceedings No. 26, 390 p.Christensen, V. and D. Pauly. 1995. Fish production, catches and the carryingcapacity of the world oceans. NAGA, the ICLARM Q. 18(3): 34-40.Christensen, V and D. Pauly. 1998.  Changes in models of aquatic ecosystemsapproaching carrying capacity.  Ecological Applications 8: S104-S109.Climo, L. 1993.  The status of cliff-nesting seabirds at St. Paul, Island Alaska in1992.  U.S. Fish and Wildl. Serv. Rep., AMNWR 93/15.  Homer, Alas. 53p.Cooney, R.T. 1981.  Bering Sea zooplankton and micronekton communities withemphasis on annual production.  p. 947-974. In Hood, D.W. and J.A.Calder (eds.)  The eastern Bering Sea shelf: Oceanography andResources, Vol. II.  Univ. Wash. Press, Seatle.  1339 p.57Dagg, M.J., J. Vidal, T.E. Whitledge, R.L. Iverson, and J.J. Goering.  1982.  Thefeeding, respiration, and excretion of zooplankton in the Bering Sea duringa spring bloom.  Deep-Sea Res. 29:45-63.Dalheim, M., A. York, J. Waite, and C. Goebel-Diaz.  1992. Abundance anddistribution of harbor porpoise (Phocoena phocoena) in Southeast Alaska,Cook Inlet and Bristol Bay. Annual report to the MMPA AssessmentProgram. Office of Protected resources, NMFS, NOAA, 1335 East-WestHighway, Silver Spring, MD 20910.Dalsgaard, J. and D. Pauly 1997. Preliminary mass-balance model of PrinceWilliam Sound, Alaska, for the pre-spill period, 1980-1989. FisheriesCentre Research Reports 5(2), 84 pp.Diamond, J. 1997. Guns, Germs and Steel: the Fates of Human Societies.Norton, NY, USA, 48pp.Dragoo, B.K. and K. Sundseth.  1993.  The statusof northern fulmars, kittiwakes, and murres at St. George Island, Alaska.U.S. Fish and Wildl. Serv. Rep., AMNWR 93/10. Homer, Alas. 92 p.Dwyer, D. A., K. M. Bailey, and P.A. Livingston. 1987.  Feeding habits and dailyration of walleye pollock (Theragra chalcogramma) in the eastern BeringSea, with special reference to cannibalism. Can. J. Fish. Aquat. Sci. 44:1972-1984.English, T.S. 1979.  Acoustic assessments of zooplankton and juvenile pollock inthe southeastern Bering Sea.  PROBES Prog. Rep. Vol. I. Dep. Oceanogr.Univ. of Washington, Seattle.Evans, P.G.H. 1987. The natural history of whales and dolphins. Facts on FilePublications, New York. 343p.Evans, S. 1984.  Energy budgets and predation impact of dominant epibenthiccarnivores on a shallow soft bottom community at the Swedish west coast.Estuar. Coast. Shelf Sci.:18:651-672.Fay, F.H., L.L. Eberhardt, B.P. Kelly, J.J. Burns and L.T. Quakenbush. 1997.Status of the Pacific walrus population, 1950-1989.  Mar. Mamm. Sci. 13:537-565.Feder, H. 1978.  Survey of the epifaunal invertebrates of the southeastern BeringSea. p. 1-126 In Environmental Assessment of the Alaskan ContinentalShelf, Annual Reports vol. 4.  U.S. Dep. Commer., Natl. Oceanic Atmos.Admin., Environ. Res. Lab. Boulder, CO.Feder, H.M. and S.C. Jewett.  1981.  Feeding interactions in the eastern BeringSea with emphasis on the benthos. p. 1229-1261.  In Hood, D.W. and J.A.Calder (eds.)  The eastern Bering Sea shelf: Oceanography andResources, Vol. II.  Univ. Wash. Press, Seatle.  1339 p.Ford, J.K.B., G.M. Ellis, L.G. Barrett-Lennard, A.B. Morton, R.S. Palm, and K.C.Balcomb III. 1998. Dietary specialization in two sympatric populations of58killer whales (Orcinus orca) in coastal British Columbai and adjacentwaters. Can. J. Zool. 76: 1456-1471.Fritz, L. W. 1996.  Squid and other species.  In: Stock assessment and fisheryevaluation report for the groundfish resources of the Bering Sea/AleutianIslands regions.  North Pacific Fishery Management Council, 605 W. 4thAve., Suite 306, Anchorage, AK 99501.Froese, R. and D. Pauly (Ed). 1998. FishBase 98. Concepts, design and datasources. ICLARM, Manila, Philippines. 290 p.Frost, K.J. 1985. The ringed seal (Phoca hispida). Alaska Dep. Fish Game,game Tech. Bull. 7: 79-87.Frost, K.J. and Lowry, L.F. 1981. Foods and trophic relationships of cetaceans inthe Bering Sea. p. 825-836. In D.W. Hood and J.A. Calder  (eds.). Theeastern Bering Sea shelf: oceanography and resources, Vol. 2, Universityof Washington Press, Seattle.Frost, K.J. and L.F. Lowry 1990. Distribution, abundance and movements ofbeluga whales, Dephinapterus leucas, in coastal waters of western Alaska.p. 39-57. In T.G. Smith, D.J. St. Aubin and J.R. Geraci (eds.), Advances inresearch on the beluga whale, Dephinapterus leucas. Can. Bull. Fish.Aquat. Sci. 224.Frost, K.J. and L.F. Lowry 1995. Distribution and abundance of beluga whales inBristol Bay, Alaska, 1993-1994. Alaska Beluga Whale Committee Rep. 95-1. 14 pp.Fujioka, J.T., M.F. Sigler, and S.A. Lowe.  1996.  Sablefish. In: Stock assessmentand fishery evaluation report for the groundfish resources of the BeringSea/Aleutian Islands regions.  North Pacific Fishery Management Council,605 W. 4th Ave., Suite 306, Anchorage, AK 99501.Fukuhara, F.M. 1985.  Biology and fishery of the southeastern Bering Sea redking crab (Paralithodes camtschatica, Tilesius).  U.S. Dep. Commer.,NWAFC Processed Report 85-11. 170p.Gaskin. D.E. 1984.  The harbour porpoise Phocoena phocoena (OL.): Regionalpopulations, status, and information on direct and indirect catches. Rep.Int. Whaling Comm. 34: 569-586.Gislason, H. 1993.  Effect of changes in recruitment levels on multispecies long-term predictions. Can. J. Fish. Aquat. Sci. 50:2315-2322.Gosho, M.E., D.W. Rice and J.M. Breiwick. 1984. The sperm whale, Physetermacrocephalus. Mar. Fish. Rev. 46:54-64.Groot, C., L. Margolis and W. C. Clarke (eds), 1995. Physiological ecology ofPacific Salmon. UBC Press, Vancouver. 510 p.59Haflinger, K.  1981.  A survey of benthic infaunal communities of thesoutheastern Bering Sea shelf.  p. 1091-1103.  In: Hood, D.W. and J.A.Calder (eds.)  The eastern Bering Sea shelf: Oceanography andResources, Vol. II.  Univ. Wash. Press, Seatle.  1339p.Haggan, N., J. Archibald, and S. Salas 1998. Knowledge gains power whenshared. Pages 8-13 in D. Pauly, T.J. Pitcher, and D. Preikshot (Eds) Backto the future: reconstructing the Strait of Georgia ecosyst m. FisheriesCentre Research Reports 6(5): 99pp.Hall, S.J. 1999. The effects of fishing on marine ecosystems. Fish Biology andAquatic Resources Series, Volume 1, Blackwell Science Oxford, UK,274pp..Hamner, W.  1983.  Gelatinous zooplankton of the Bering Sea.  p. 213-229. InPROBES, Processes and Resources of the Bering Sea Shelf. Final Report- Vol. II.Hatch, S. A., G. V. Byrd, D.B. Irons, and G.L. Hunt Jr.  1993.  Status and ecologyof kittiwakes (Rissa tridactyla nd R. brevirostris) in the North Pacific.  p.140-153.  In K. Vermeer, K.T. Briggs, K.H. Morgan, and D. Siegel-Causey(eds.)  The status, ecology, and conservation of marine birds of the NorthPacific.  Can. Wildl. Serv. Spec. Publ., Ottawa.Hill, P.S., D.P. DeMaster, and R.J. Small. 1997. Alaska marine mammalpopulation assessments, 1996.  NOAA Tech. Memo. NMFS-AFSC-78.Hobbs, R.C., and J.A. Lerczak. 1993. Abundance of Pacific white-sided dolphinand Dall’s porpoise in Alaska estimated from sightings in the North PacificOcean and the Bering Sea during 1987 through 1991.  Annual report to theMMPA Assessment Program, Office of Protected Resources, NMFS,NOAA, 1335 east-WestHolling, C.S. 1973. Resilience and stability of ecological systems. Ann. Rev.Ecol. Syst. 4:1-23.Hollowed, A, S. R. Hare, and W.S. Wooster.  1998.  Pacific basin climatevariability and patterns of northeast Pacific marine fish production.  In: AhaHuliko’a Workshop ProceedingsHunt, G.L. Jr., B. Burgeson, and G.A. Sanger.  1981.  Feeding ecology ofseabirds of the eastern Bering Sea.  p. 629-647.  In Hood, D.W. and J.A.Calder (eds.)  The eastern Bering Sea shelf: Oceanography andResources, Vol. II.  Univ. Wash. Press, Seatle.  1339 p.Ianelli, J.N., T.K. Wilderbuer, and T.M. Sample.  1996.  Stock assessment ofGreenland turbot.  In: Stock assessment and fishery evaluation report forthe groundfish resources of the Bering Sea/Aleutian Islands regions.  NorthPacific Fishery Management Council, 605 W. 4th Ave., Suite 306,Anchorage, AK 99501.60Innes, S., D.M. Lavigne, W.M. Earle and K.M. Kovacs. 1987. Feeding rates ofseals and whales.  J. Anim. Ecol. 56: 115-130.IPHC [International Pacific Halibut Commission].  1996.  Pacific halibutpopulation assessment and fishery evaluation.  In: Stock assessment andfishery evaluation report for the groundfish resources of the BeringSea/Aleutian Islands regions.  North Pacific Fishery Management Council,605 W. 4th Ave., Suite 306, Anchorage, AK 99501.Ito, D.  1996.  Other rockfish.  In: Stock assessment and fishery evaluation reportfor the groundfish resources of the Bering Sea/Aleutian Islands regions.North Pacific Fishery Management Council, 605 W. 4th Ave., Suite 306,Anchorage, AK 99501.Ito, D. and J. Ianelli.  1996.  Pacific ocean perch.  In: Stock assessment andfishery evaluation report for the groundfish resources of the BeringSea/Aleutian Islands regions.  North Pacific Fishery Management Council,605 W. 4th Ave., Suite 306, Anchorage, AK 99501.Ivashin, M.V., and L.M. Votrogov. 1981. Minke whales, Balaenopteraacutorostrata dividsoni, inhabiting inshore waters of the Chukotka coast.Rep. Int. Whaling Comm. 31: 231.Iverson, R.L. and J.J. Goering.  1979.  Primary production and phytoplanktoncomposition in the southeast Bering Sea.  p. 144-162.  In: Progress Reportof Processes and Resources of the Bering Sea Shelf, Vol. 1.  Univ. Alaska,Fairbanks.Jefferson, T.A., S. Leatherwood, and M.A. Webber. 1993. FAO speciesidentification guide. Marine mammals of the world.  United NationsEnvironmental Program, Food and Agriculture Organization of the UnitedNations, Rome.Johnson, M.L., C.H. Fiscus, B.T. Ostenson and M.L. Barbour. 1966. Marinemammals. p. 877-924. In N.J. Wilimovsky and J.N. Wolfe (eds),Environment of the Cape Thompson region, Alaska, U.S. Atomic EnergyComm. (available from Natl. Tech. Inf. Serv., Springfield, VA, as PNE-481).Kasamatsu, F. and T. Hata 1985. Notes on minke whales in the Okhotsk Sea-west Pacific area. Rep. Int. Whaling Comm. 35: 299-304.Kasuya, T. and S. Ohsumi 1984. Further analysis of the Baird’s beaked whalesoff the Pacific coast of Japan Scu reo Wakes Res, Inst. 37: 61-83.Kasuya, T. and T. Miyashita. 1988. Distribution of sperm whale populations in thenorth Pacific.  Sci. Rep. Whales Res. Inst. 39: 31-75.Kawakami, T. 1980. A review of sperm whale food. Sci. Rep. Whales Res. Inst.Tokyo 32: 199-218.61Kawamura, A. 1980. A review of food of balaenopterid whales. Sci. Rep. WhalesRes. Inst. Tokyo 32: 155-197.Kelly, B.P. 1988. Ringed seal, Phoca hispida. p. 57-75. In J.W. Lentfer (ed.),selected marine mammals of Alaska. Species accounts with research andmanagement recommendations. Marine Mammal Commission,Washington D.C.Kleinenberg, S.W., A.V. Yablokov, B.M. Del’kovich, and M.N. Tarasevich. 1964.Belukha: Opyt monograficheskogo issledovaniya vida. [Beluga(Delphinapterus leucas): Investigation of the species.] [In Russ.] Izd. Akad.Nauk SSSR, Moscow, Natl. Tech. Inf. Serv, Springfield, VA, as TT67-51345).Lander  R.H. 1980. Summary of northern fur seal data and collection procedures.Vol. 1: Land data of the United States and Soviet Union (excluding tag andrecovery records). U.S. Dep. Commer., NOAA Tech. Memo. NMFSF/NWC-3, 315 p.Lander, R.H. 1981. A life table and biomass estimate for Alaskan fur seals. Fish.Res. 1:55-70.Livingston, P.A. (editor) 1991. Groundfish food habits and predation oncommercially important prey species in the eastern Bering Sea from 1984to 1986.  U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-207.240p.Livingston, P.A.  1993.  Importance of predation by groundfish, marine mammalsand birds on walleye pollock Theragra chalcogramma and Pacific herringClupea pallasi in the eastern Bering Sea.  Mar. Ecol. Prog. Ser. 102:205-215.Livingston, P. A. and B. J. Goiney, Jr.  1983.  Food habits literature of NorthPacific marine fishes:  A review and selected bibliography.  U.S. Dept.Commerce, NOAA Tech. Memo. NMFS F/NWC-54.Livingston, P.A. and J. Jurado-Molina. In press.  A multispecies virtual populationanalysis of the eastern Bering Sea.  ICES J. Mar. Sci.Livingston, P.A. and R.D. Methot.  1998. Incorporation of predation into apopulation assessment model of eastern Bering Sea walleye pollock.  p.x-x. In Fishery Stock Assessment Models of the 21st Century.  Alaska SeaGrant College Program Publication.  AK-SG-98-01.Livingston, P.A., A. Ward, G.M. Lang, and M-S. Yang.  1993.  Groundfish foodhabits and predation on commercially important prey species in theeastern Bering Sea from 1987 to 1989.  U.S. Dep. Commer., NOAA Tech.Memo. NMFS-AFSC-11.  192p.Lockyer, C. 1976. Body weights of some species of large whales. J. Cons. Int.Explor. Mer 36: 259-273.62Lockyer, C. 1981a. Estimates of growth and energy budget for the sperm whale,Physeter catodon, FAO Fish. Ser. (5), Mammals in the Seas 3, 489-504.Lockyer, C. 1981b. Growth and energy budgets of large baleen whales from thesouthern hemisphere. p. 379-488. In FAO Advisory Committee in MarineResources Research Working Party on Marine Mammals in the seas.Volume 3. General papers and large cetaceans.Loughlin, T.R. 1998. The Steller sea lion: a declining species.  BiosphereConservation 1: 91-98.Loughlin, T.R. and M.A. Perez. 1985. Mesoplodon stejnegeri. Mamm. Species250: 1-6.Lowry, L.F. 1985. The spotted seal (Phoca largha). Alaska Dep. Fish Game,Game Tech. Bull. 7:89-96.Lowry L.F. and K.J. Frost. 1981. Feeding and trophic relationships of Phocidseals and walruses in the eastern Bering Sea. p. 813-824.  In D.W. Hoodand J.A. Calder (eds.), The eastern Bering Sea shelf: Oceanography andresources, Vol. 2, Univ. Washington Press, Seattle.Lowry, L.F. and K.J. Frost. 1984. Foods and feeding of bowhead whales inwestern northern Alaska. Sci. Rep. Whales Res. Inst. Tokyo 35: 1-16.Lowry, L.F., K.J. Frost, D.G. Calkins, G.L. Swartzman, and S. Hills. 1982.Feeding habits, food requirements and status of Bering Sea marinemammals. Counc. Doc. Nos. 19 and 19A (annotated bibliography), 574 p.North Pacific Fisheries Management Council, P.O. Box 103136,Anchorage AK 99510.Lowry, L.F., K.J. Frost, and T.R. Loughlin. 1989. Importance of walleye pollock inthe diets of marine mammals in the Gulf of Alaska and Bering Sea andimplications for fishery management.  p. 701-726. In Proceedings of theInternational Symposium on the Biology and Management of WalleyePollock. Alaska Sea Grant Report 89-1, Univ. Alaska, Fairbanks, AK.Lucas, F.A. 1899. The causes of mortality among seals. p. 75-98 In The fur sealsand fur seal islands of the North Pacific Ocean. Part 3. Washington DC:U.S. Treasury Department.Mauchline, J.  1980.  The biology of mysids and euphausiids. p. 1-681. Vol.18.In: Blaxter, J.H.S., F.S. Russell and M. Yonge (eds). Adv. in Mar. Biol.Academic Press, London.McDonald, J., H.M. Feder, and M. Hoberg.  1981.  Bivalve molluscs of thesoutheastern Bering  Sea.  p. 1155-1204. In: Hood, D.W. and J.A. Calder(eds.)  The eastern Bering Sea shelf: Oceanography and Resources, Vol.II.  Univ. Wash. Press, Seatle.  1339 p.63Merrick, R. 1995. The relationship of the foraging ecology of Steller sea lions(Eumetopias jubatus) to their population decline in Alaska. PhD Thesis,University of Washington.Merrick, R.L., M.K. Chumbley, and G.V. Byrd. 1997. Diet diversity of Steller sealions (Eumetopias jubatus) and their population decline in Alaska: apotential relationship. Canadian Journal of Fisheries and Aquatic Science54: 1342-1348.Meshcheryakova, I. M. 1964 (translated 1968). The quantitative distribution ofplankton in the southeastern Bering Sea in the summers of 1958 and1959. p. 147-158. In P. A. Moiseev (eds.), Soviet fisheries investigations inthe northeastern Pacific, Part II. Pishchevaya Promyshlennost, VNIROProc. Vol. 49 and TINRO Proc. Vol. 51, Moscow.Meshcheryakova, I. M. 1970. (translated 1972). Winter and spring plankton in thesoutheastern part of the Bering Sea. p. 98-114 In P. A. Moiseev (eds.),Soviet fisheries investigations in the northeastern Pacific. PishchevayaPromyshlennost, VNIRO Proc. Vol. 70 and TINRO Proc. Vol. 72, Moscow.Motoda, S. and T. Minoda.  1974.  Plankton of the Bering Sea. p. 207-241.  InHood, D.W. and E.J. Kelley (eds.) Oceanography of the Bering Sea.  Inst.Mar. Sci., Occ. Pub. No. 2.  Univ. Alaska, Fairbanks.NRC [National Research Council]. 1996. The Bering Sea ecosystem. NationalAcademy Press, Washington, D.C. 207 pp.Neiman, A.A.  1968.  Quantitative distribution of benthos on the shelf and uppercontinental slope in the eastern part of the Bering Sea.  p.143-217. In: P.A.Moiseev (editor), Soviet Fisheries Investigations in theNemoto, T. 1957. Foods of baleen whales in the northern Pacific. Sci. Rep.Whales Res. Inst. Tokyo 12: 33-89.Nemoto, T. 1959. Food of baleen whales with reference to whale movements.Sci. Rep. Whales Res. Inst. Tokyo 14: 144-290.Nemoto, T. 1970. Feeding pattern of baleen whales in the ocean. p. 241-252.  InJ.H. Steele (ed.), marine food chains, Oliver and Boyd, Edinburgh, UK.Niebauer, H.J.  1988.  Effects of El Nino-Southern Oscillation and North Pacificweather patterns on interannual variability in the subarctic Bering Sea.  J.Geophys. Res. 93:5051-5068.NMFS [National Marine Fisheries Service].  1988.  Marine Mammal ProtectionAct of 1972 Annual Report 1987-88.  U.S. Dep. Comm., National Oceanicand Atmospheric Admin., National Marine Fisheries Service, Office ofProtected Resources, Silver Spring, MD 20910.NMFS [National Marine Fisheries Service]. 1991. Recovery Plan for the NorthernRight Whale (Eubalaena glacialis). Prepared by the Right Whale Recovery64Team for the National Marine Fisheries Service.  Silver Spring, Maryland.86 p.NPFMC [North Pacific Fishery Management Council].  1996.  Stock assessmentand fishery evaluation report for the groundfish resources of the BeringSea/Aleutian Islands regions.  North Pacific Fishery Management Council,605 West 4th Ave., Suite 306, Anchorage, AK 99501.Novikov, N.P. 1970.  Biologiia maloglazogo dolgokhovosta Chalinura pectoralis vsevernoi chasti Tikhogo okeana (Biology of Chalinura pectoralis in theNorth Pacific).  Tr. Vses. Nauchno-Issled. Inst. Morsk. Rybn. Khoz.Okeanogr. 70 (Izv. Tikhookean. Nauchno-Issled. Inst. Rybn. Khoz.Okeanogr. 72):300-326.  In Russ.  (Transl. by Isr. Program Sci. Transl.1972, 9. 304-331 in P.A. Moiseev (editor), Soviet Fisheries Investigationsin the Northeastern Pacific, Part 5, available U.S. Dep. Commer., Natl.Tech. Inf. Serv., Springfield, Va., as TT-71-50127.)Odum, E.P. 1969. The strategy of ecosystem development. Science 104:262-270.Odum, E.P. 1971. Fundamentals of ecology.  W.B. Saunders Co., Philadelphia.574 p.Ohsumi, S. 1966. Sexual segregation of the sperm whale in the North Pacific.Sci. Rep. Whales Res. Inst. Tokyo 20: 1-16.Ohsumi, S. 1979. Interspecies relationships among some biological parametersin cetaceans and estimation of the natural mortality coefficient of thesouthern hemisphere minke whale. Rep. Int. Whaling Comm. No. 29, pp.397-406.Ohsumi, S., Y. Shimadzu, and T. Doi 1971. The seventh memorandum on theresults of Japanese stock assessment of whales in the North Pacific. Rep.Int. Whaling Comm. 21: 76-89.Ohsumi, S., and S. Wada. 1974. Status of whale populations in the North Pacific.1972. Rep. Int. Whal. Comm. 24: 114-126.Okey, T. and D. Pauly (eds)  1998.  Trophic mass balance model of Alaska'sPrince William Sound ecosystem, for the post-spill period 1994-1996.Fisheries Centre Research Reports 6(4), 144pp.Okutani, T., and T.Nemoto. 1964. Squids as the food of sperm whales in the Bering Sea andAlaskan Gulf. Sci. Rep. Whales Res. Inst. Tokyo 18:111-122.Opitz, S. 1993. A quantitative model of trophic intereactions in a Caribbean CoralReef Ecosystem. p. 259-267. In V. Christensen and D. Pauly (eds.)Trophic models of aquatic ecosystems. ICLARM Conf. Proc. 26, 390 p.Otto, R.S.  1986.  Management and assessment of eastern Bering Sea king crabpopulations, p. 83-106.  In G.S. Jamieson and N. Bourne (eds.) North65Pacific Workshop on Stock Assessment and Management ofInvertebrates.  Can. Spec. Publ. Fish. Aquat. Sci. 92.Otto, R.S., J.A. Haaga, and R.A. MacIntosh.  1997.  Report to industry on the1996 eastern Bering Sea crab survey.  U.S. Dep. Commer., Natl. Oceanicand Atmos. Admin., Alaska Fisheries Science Center AFSC ProcessedReport 97-02. 52p.Paul, A.J. and A. Fuji.  1989.  Bioenergetics of the Alaskan crab Chionoecetesbairdi (Decapoda: Majidae).  J. Crust. Biol. 9:25-36.Pauly, D.  (ed.) 1998. The use of Ecopath with Ecosim to evaluate strategies forsustainable exploitation of multi-species resources. Fisheries CentreResearch Reports 6 (2),  49pp.Pauly, D. and V. Christensen. 1995. Primary production required to sustainglobal fisheries. Nature 374:255-257.Pauly, D. and V. Christensen 1996. Mass-balance models of North-easternPacific ecosystems. Fisheri s Centre Research Reports 4(1),  131pp.Pauly, D., T.J. Pitcher,  and D. Preikshot (eds) 1998a. Back to the future:reconstructing the Strait of Georgia ecosystem. Fisheries Centre ResearchReports 6(5), 99pp.Pauly D., V. Christensen, J. Dalsgaard, R. Froese, F. Torres Jr. 1998b. Fishingdown marine food webs. Science 279: 860-863.Pauly, D., A.W. Trites, E. Capuli and V. Christensen. 1998c. Diet compositionand trophic levels of marine mammals.  ICES Journal of Marine Science55: 467-481.Pearson, W.H., D.L. Woodruff, and B.J. Higgins. 1984.  Feeding ecology ofjuvenile king and Tanner crab in the southeastern Bering Sea.  FinalReport, Outer Continental Shelf Environmental Assessment Program.Battell Pacific Northwest Lab, Richland, WA 99352Perez, M.A. 1990. Review of marine mammal and prey population information forBering Sea ecosystem studies. NOAA technical memorandum NMFSF/NWC-186.Perez, M.A. and M.A. Bigg. 1981. An assessment of the feeding habits of thenorthern fur seal in the eastern north Pacific Ocean and eastern BeringSea. 140 p (unpubl. Report, available from the National Marine MammalLaboratory, Sand Point Way, Seattle WA)Perez, M.A. and M.A. Bigg. 1986. Diet of northern fur seals (Callorhinus ursinus)off western north America. Fishery Bulletin 84:957-971.Pitcher, K.W. 1990. Major decline in number of harbor seals, Phoca vitulinarichardsi) on Tugidak Island, Gulf of Alaska. Mar. Mamm. Sci. 6: 121-134.66Pitcher, T.J. (ed.) 1996. Reinventing fisheries management. Fisheries CentreResearch Reports 4(2), 84 pp. Pitcher, T.J. 1998a. Pleistocene pastures:Steller’s sea cow and sea otters in the Strait of Georgia. Pages 49-52 inPauly, D., Pitcher, T.J. and Preikshot, D. (Eds) Back to the future:reconstructing the Strait of Georgia ecosystem. Fisheries Centre ResearchReports 6(5): 99pp.Pitcher, T.J. 1998b. Back To the future: a novel methodology and policy goal infisheries. Pages 4-7 in D. Pauly, T.J. Pitcher, and D. Preikshot (Eds), Backto the future: reconstructing the Strait of Georgia ecosystem. FisheriesCentre Research Reports 6(5): 99pp.Pitcher, T.J., D. Pauly, N. Haggan, and D. Preikshot. 1999.  Back to the future: amethod employing ecosystem modelling to maximise the sustainablebenefits from fisheries. Alaska Sea Grant Symposium 1998 (inpress).Polovina, J.J. 1984. Model of a coral reef ecosystem. Part I. TheEcopath application and its application to French Frigate Shoals. CoralReefs 3:1-11.Popov, L.A. 1976. Status of main ice forms of seal inhabiting waters of theU.S.S.R. and Adjacent to the country marine areas. BerginACMRR/MM/SC/51. 18 p.Queirolo, L.E., L.W. Fritz, P.A. Livingston, M.R. Loefflad, D.A. Colpo, and Y.L.deReynier.  1995.  Bycatch, utilization, and discards in the commercialgroundfish fisheries of the Gulf of Alaska, eastern Bering Sea, and AleutianIslands.  NOAA Tech. Memo. NMFS-AFSC-58. 148p.Quinn, T.J., II, and H.J. Neibauer.  1995.  Relation of eastern Bering Sea walleyepollock (Theragra chalcogramma) recruitment to environmental andoceanographic variable, p. 497-507.  In R.J. Beamish (ed.) Climate changeand northern fish populations.  Can. Spec. Publ. Fish. Aquat. Sci. 121.Radchenko, V.I.  1992.  The role of squids in the pelagic ecosystem of the BeringSea.  Oceanology 32(6):762-767.Reilly, S.B., and J. Barlow. 1986. Rates of increase in dolphin population size.Fishery Bulletin. 84: 527-533.Rice, D.W. 1968. Stomach contents and feeding behaviour of killer whales in theeastern North Pacific. Nor. Hvalfangst-Tid. 57: 35-38.Rice, D.W. 1978. The humpback whale in the North Pacific: distribution,exploitation and numbers. Appendix 4. Pp. 29-44. In K.S. Norris and R.R.Reeves (eds.). Report on a workshop on problems related to humpackwhales (megaptera novaeangliae) in Hawaii. U.S. Dep. Commer., Nat.Tech. Info. Serv. PB-280 794. Springfield VA.Rice, D.W. 1986. Beaked whales. p. 102-109. In D. Haley (ed.). Marinemammals of eastern North Pacific and arctic waters, 2nd Ed., rev., PacificSearch Press, Seattle, WA.67Rice, D.W. and A.A. Wolman. 1971. The life history and ecology of the graywhale, Eschrichtius robustus. Am. Soc. Mamm. Spec. Publ. 3. 142 p.Rice, R.L. 1981.  Feeding habits of crangonid shrimps and some aspects ofsediment-detrital food systems in lower Cook Inlet, Alaska. NOAA/OMPAPrincipal Investigation Report on the Environmental Assessment of theAlaskan Continental Shelf, Vol. 14, Biological Studies, p.611-676.Ricker, W.E. 1975. Computation and interpretation of biological statistics of fishpopulations. Bull. Fish. Res. Board Can. 191. 382 p.Schneider, D. and G.L. Hunt.  1982.  Carbon flux to seabirds in waters withdifferent mixing regimes in the southeastern Bering Sea.  Mar. Biol.67:337-344.Seaman, G.A., K.J. Frost, and L.F. Lowry. 1985. Investigations of belukhawhales in coastal waters of western and northern Alaska. Part I.Distribution, abundance and movements. U.S. Dep. Commer., NOAA,OCSEAP Final Rep. 56: 153-220.Sease, J.L. 1992. Status review: harbor seals (Phoca vitulina) in Alaska. NMFSAFSC Processed Rep. 92-15. 74 pp.Seitz, R.D. and L.C. Schaffner.  1995.  Population ecology and secondaryproduction of the polychaete Loimia medusa (Terebellidae).  Mar. Biol.121:701-711.Shuntov, V.P. 1972.  Seasonal distribution of black and arrow-toothed halibuts inthe Bering Sea. p. 397-408 In P.A. Moiseev (ed) Soviet FisheriesInvestigations in the Northeastern Pacific, Part V.  (Translated fromRussian by Israel Program for Scientific Translations for U.S. Dept.Commerce, National Marine Fisheries Service, NTIS doc. TT 71-50127.)Sinclair, E., T. Loughlin, and W. Pearchy. 1994. Prey selection by northern furseals (Callorhinus ursinus) in the eastern Bering Sea. Fish. Bull. 92: 144-156.Sissenwine, M.P., E.B Cohen, and M.D. Grosslein.  1984.  Structure of theGeorges Bank ecosystem.  Rapp. P.-v. Cons. int. Explor. Mer 183:243-254.Sleptsov, M.M. 1961. Observations of small cetaceans in far eastern seas andNW Pacific. [in Russian]. TR. Inst. Morfol. Zhivotnykh Akad. Nauk SSSR34: 136-143. (Transl. By M. Slessers, U.S. Nav. Oceanogr. Off.,Washington, D.C., 1970. NOO Transl. 482, 11p.)Small, R.J. and D. DeMaster. 1995. Alaska marine mammal populationassessments 1995. US Dept. Commer., NOAA Tech Memo. NMFS-AFSC-57. 93 p.68Somerton, D. 1981.  Life history and population dynamics of two species ofTanner crab, Chionoecetes bairdi and C. opilio, in the eastern Bering Seawith implications for the management of the commercial harvest.  Ph.D.thesis, University of Washington, Seattle, WA. 220p.Spencer, P.D. and J.S. Collie. 1996.   A simple predator-prey model of exploitedmarine fish populations incorporating alternative prey.  ICES J. Mar. Sci.53:615-628.Springer, A.M., C.P. McRoy, and M.V. Flint.  1996.  The Bering Sea Green Belt:shelf-edge processes and ecosystem production.  Fish. Oceanogr. 5:205-223.Steele, J.H. and E.W. Henderson. 1984. Modeling long-term fluctuations in fishstocks. Science 224:985-987.Sugimoto, T. and K. Tadokoro.  1997.  Interannual-interdecadal variations inzooplankton biomass, chlorophyll concentration and physical environmentin the subarctic Pacific and Bering Sea.  Fish. Oceanogr. 6:74-93.Tarasevich, M.N. 1968. Food connections of the sperm whales of the NorthernPacific. [in Russian]. Zool. Zhur. 47: 595-601. (Transl. By K. Coyle, Univ.Alaska, Fairbanks, 1982, 11 p.).Thompson, G.G. and M.W. Dorn.  1996.  Pacific cod.  In: Stock assessment andfishery evaluation report for the groundfish resources of the BeringSea/Aleutian Islands regions.  North Pacific Fishery Management Council,605 W. 4th Ave., Suite 306, Anchorage, AK 99501.Thornsteinson, F.V. and C.J. Lensink. 1962. Biological observations of Stellersea lions taken during an experimental harvest. J. Wild. Manage. 26:353-359.Tomilin, A. 1957. Mammals of the USSR and Adjacent Countries, p. In V.G.Heptmer (ed.). Cetacea: vol. IX, Nauk USSR.Trites, A.W. 1989. Estimating the juvenile survival rate of male northern fur seals(Callorhinus ursinus). Can. J. Fish. Aquat. Sci. 46: 1428-1436.Trites, A.W. 1991. Fetal growth of northern fur seals: life history strategy andsources of variation. Can. J. Zool. 69: 2608-2617.Trites, A.W. and M.A. Bigg 1996. Physical growth of northern fur seals: seasonalfluctuations and migratory influences.   J. Zool. (Lond.).  238: 459-4 2.Trites, A.W. and P.A. Larkin. 1989. The decline and fall of the Pribilof fur seal(Callorhinus ursinus): a simulation study.  Can. J. Fish. Aquat. Sci.46:1437-1445.69Trites, A.W., and P.A. Larkin. 1996. Changes in the abundance of Steller sealions (Eumetopias jubatus) in Alaska from 1956 to 1992: how many werethere? Aquat Mamm. 22:153-166.Trites, A.W. and P.A. Larkin. 1992. The status of Steller sea lion populationsand the development of fisheries in the Gulf of Alaska and AleutianIslands.  A report of the Pacific States Marine Fisheries Commissionpursuant to NOAA award no. NA17FD0177.  134 p.Trites, A.W., P.A. Livingston, M.C. Vasconcellos, S. Mackinson, A.M. Springerand D. Pauly. In press. Ecosystem Considerations and the Limitations ofEcosystem Models in Fisheries Management: Insights from the BeringSea.  In Proceedings of Ecosystem Considerations in FisheriesManagement.  16th Lowell Wakefield Fisheries Symposium and AmericanFisheries Society joint meeting.  Anchorage, Alaska, USA.  September 30– October 3, 1998.Trites, A.W. and D. Pauly. 1998. Estimating mean body masses of marinemammals from maximum body lengths.  Can. J. Zool. 76: 886-896.Turnock, B.J., and T.J. Quinn. 1991. The effect of responsive movement onabundance estimation using line transect sampling. Biometrics 47: 701-715.Ulanowicz, R.E. 1986. Growth and development; ecosystem phenomology.Springer Verlag, New York. 203 p.Vasconcellos, M., S. Mackinson, K. Sloman and D. Pauly. 1997. The stability oftrophic mass-balance models of marine ecosystems: a comparativeanalysis. Ecological Modelling 100:125-134.Venrick, E.L., J.A. McGowan, D.R. Vayan, and T.L. Hayward. 1987. Climate andchlorophyll a: Long-term trends in the central North Pacific Ocean. Science238: 70-72.Vitousek, P.M., P.R. Ehrlich, A.H. Ehrlich and P.A. Matson. 1986. Humanappropriation of the products of photosynthesis. Bioscience 36 368-373.Wada, S. 1973. The ninth memorandum on the population assessment ofwhales in the North pacific. Rep. Int. Whal. Comm. 23: 164-169.Wada, S. 1980. Japanese whaling and whale sighting in the North Pacific 1978season. Rep. Int. Whaling Comm. 30: 415-424.Wada, S. 1981. Japanese whaling and whale sighting in the North Pacific 1979season. Rep. Int. Whaling Comm. 31: 783-792.Walsh, J.J., and C.P. McRoy. 1986. Ecosystem analysis in the southeasternBering Sea. Cont. Shelf Res. 5: 259-288.70Walters, C., V. Christensen, D. Pauly. 1997. Structuring dynamic models ofexploited ecosystems from trophic mass-balance assessments. Rev. FishBiol. Fish. 7, 139-172.Walters, G.E. and T.K. Wilderbuer.  1996.  Flathead sole.  In: Stock assessmentand fishery evaluation report for the groundfish resources of the BeringSea/Aleutian Islands regions.  North Pacific Fishery Management Council,605 W. 4th Ave., Suite 306, Anchorage, AK 99501.Walters, G.E., K. Teshima, J.J. Traynor, R.G. Bakkala, J.A. Sassano,K.L.Halliday, W.A. Karp, K. Mito, N.J. Williamson, and D. M. Smith.  1988.Distribution, abundance, and biological characteristics of groundfish in theeastern Bering Sea based on results of the U.S.-Japan triennial bottomtrawl and hydroacoustic surveys during May-September 1985.  U.S. Dep.Commer., NOAA Tech. Memo. NMFS F/NWC-154.  400p.Warwick, R.M.  1980.  Population dynamics and secondary production ofbenthos.  p. 1-24. In Tenore, K.R. and B.C. Coull (eds).  Marine BenthicDynamics.  The Belle Branch Library in Marine Science Number 11.  451p. Univ. S. Carolina Press, Columbia.Wespestad, V.G.  1991.  Pacific herring population dynamics, early life history,and recruitment variation relative to eastern Bering Sea oceanographicfactors.  Ph.D. thesis, Univ. Washington, Seattle, WA, 237 p.Wespestad, V.G., J. Ianelli, L. Fritz, T. Honkalehto, G. Walters.  1996.  Walleyepollock.  In: Stock assessment and fishery evaluation report for thegroundfish resources of the Bering Sea/Aleutian Islands regions.  NorthPacific Fishery Management Council, 605 W. 4th Ave., Suite 306,Anchorage, AK 99501.Wilderbuer, T.K. 1996. Yellowfin sole. In: Stock assessment and fisheryevaluation report for the groundfish resources of the Bering Sea/AleutianIslands regions.  North Pacific Fishery Management Council, 605 W. 4thAve., Suite 306, Anchorage, AK 99501.Wilderbuer, T.K. and G.E. Walters.  1996.  Other flatfish.  In: Stock assessmentand fishery evaluation report for the groundfish resources of the BeringSea/Aleutian Islands regions.  North Pacific Fishery Management Council,605 W. 4th Ave., Suite 306, Anchorage, AK 99501.Wilderbuer, T.K. and T.M. Sample.  1996.  Arrowtooth flounder.  In: Stockassessment and fishery evaluation report for the groundfish resources ofthe Bering Sea/Aleutian Islands regions.  North Pacific FisheryManagement Council, 605 W. 4th Ave., Suite 306, Anchorage, AK 99501.Williamson, and D.M. Smith.  1988.  Distribution, abundance, and biologicalcharacteristics of groundfish in the eastern Bering Sea based on results ofthe U.S.-Japan triennial bottom trawl and hydroacoustic surveys duringMay-September 1985.  U.S. Dep. Commer., NOAA Tech. Memo NMFSF/NWC-154.  400p.71Withrow, D.E., and T.R. Loughlin 1996. Abundance and distribution of harborseals (Phoca vitulina richardsi) along the north side of the AlaskaPeninsula and Bristol Bay during 1995.  Annual report of the MMPAAssessment Program, Office of Protected Resources, NMFS, NOAA, 1335East-West Highway, Silver Spring, MD 20910.Zeh, J.E., J.C. George, and R. Suydam. 1994. Rate of increase, 1978-1993, ofbowhead whales, Balaena mysticetus. Rep. Int. Whal. Comm. 45: 339-344.Zeh, J.E., A.E. Raftery and A.A. Schaffner. 1995. Revised estimates of bowheadpopulation size and rate of increase. Unpubl. report submitted to Int. Whal.Comm. (SC/47/AS/10). 26 p.72 Appendix 1 – Mass-Balance Model DetailsECOPATH - steady state mass-balance ecosystem modelEcopath is a steady state model based on a set of simultaneous linearequations (one for each group i in the system (Christensen and Pauly1992b). The master equation simply states that at equilibrium, for all i:Production by (i) utilized within the system -catches of (i) - consumption of(i) by its predators = 0This can also be put as:å=××× ---=kjijiiiiii QBMBFEEBPB10)/(0 (1)where; Bi is the biomass of i during the period in question; P/Bi theproduction to biomass rate of i, equal to the total mortality rate (Z) underthe assumption of equilibrium (Allen 1971); EE is the ecotrophic efficiency,i.e., the fraction of the production ( / )P B P Bi i= ×  that is consumed withinthe system; F is the fishing mortality on i; M0 is the mortality rate notaccounted for by consumption within the system; Qij is the amount of iconsumed by jEcopath solves the set of simultaneous equations to produce a balancedbox model ecosystem in which the energy flows are quantified.ECOSIM - dynamic mass-balance approach for ecosystem simulationBy converting the linear equations of Ecopath models to differentialequations, Ecosim provides a dynamic mass-balance approach, suitablefor simulation (Walters et. al. 1997). Constructing a dynamic model fromequation (1) there are three changes viz; (a) replace the left side with arate of change of biomass; (b) for primary producers, provide a functionalrelationship to predict changes in (P/Bi) with biomass Bi (representingcompetition for light, nutrients and space); and (c) replace the static pool-pool consumption rates with functional relationships predicting howconsumption will change with changes in biomass of Bi and j..73Generalizing for both equilibrium and non equilibrium situations, gives:dB dt h B M B FB c B Bi o i i ij i jjn/ ( ) ( )= - - - ×=å1(2)where h(B) is a function of Bi if i is a primary producer orh B g c B Bi ij i jjn( ) ( )= ×=å1if i is a consumer, and c B Bij i j( )×  is the function usedto predict Qij from Bi and Bj (Walters et. al. 1997). For primary producers asimple saturating production relationship is used.Using previously constructed Ecopath models, Ecosim calculatescorresponding changes in biomass of each component when the fishingmortality of any particular group is altered. These dynamic simulations areplotted as coloured biomass curves. The scale differs for each curve. Byaltering the rate of flow between vulnerable and non-vulnerable preydifferent functional relationships for predators and prey can be considered.These can range from pure donor control, where the prey availabilitygoverns interactions, to top-down control where predation pressuredominates. Using equilibrium simulations, where equilibrium biomass isplotted over a range of F values, Ecosim provides the facility to predict thepotential equilibrium yield for the fished group.74Appendix 2 – Species Assemblage DetailsMammals and BirdsOver 20 species of marine mammals feed on the eastern continental slope orshelf (NRC 1996).  The majority of these are most abundant or occur solelyduring the summer months, May-October. Those most abundant in winter(November-April) include: bowhead whales, Dall’s porpoise, Pacific walrus, ringedseals and bearded seals.  Species that were not deemed to feed in theshelf/slope region of the eastern Bering Sea (but may feed in other areas of thenorth Pacific) were not included in the model. They include: Sei, Right, Cuviersand Bairds whales.  The remaining 21 species of marine mammals that feed inthe eastern Bering Sea were aggregated into 7 groups.These 7 groups of marine mammals included: 1) baleen whales (fin whales,minke whales, blue whales, humpback whales, bowhead whales, and graywhales); 2) sperm whales; 3) toothed whales (beluga whales, killer whales, Dall’sporpoise, and harbor porpoise); 4) beaked whales, 5) walrus and bearded seals;6) Steller sea lions; and 7) seals (northern fur seals, harbor seals, spotted seals,ribbon seals, and ringed seals).Sperm whales were considered a separate group from the toothed whales for tworeasons.  First, their high biomass and specialized diet would heavily bias thetoothed whale group.  Second, sperm whales were exploited at much higher ratesthan other toothed whales.Pinnipeds were split into three groups. In particular, Pacific walruses and beardedseals were not included with other pinnipeds because their high biomass andpeculiar diets would have heavily biased the pinniped group.  Second, we wereparticularly interested in understanding the changes that have occurred to Stellersea lions; a species that declined dramatically over the past two decades andwas recently declared an endangered species (Loughlin 1998).Abundance.  Population estimates for the 1979-1985 period came from Perez(1990), the NRC report (1996), and the 1987-1988 Marine Mammal ProtectionAct annual report (NMFS 1988). Estimates in the latter two documents mostlyconsider populations for the whole of the North Pacific.  We assumed thatportions of these populations are present in the Bering Sea during some part ofthe summer.   Population estimates used in the model are contained in TableA2.1.Growth rate.  The maximum rate of population growth rate for northern fur sealsand other pinnipeds is believed to be about 12% per year (Small and DeMaster1995). The P/B ratio was therefore set at 6%, half of the maximum. Maximumrate of population increase for whales is 4% (Reilly and Barlow 1986) and the P/Bratio was estimated to be 2% (half of rmax).75Table A2.1. Estimated numbers of marine mammals and total biomass duringsummer (May-Oct) and winter (Nov-Dec) in the 1950s and 1980s.  The seasonalestimates were averaged to determine annual biomass.  Details are contained in thespecies descriptions.May-OctNov-AprMay-OctNov-Apr May-OctNov-AprMay-OctNov-AprBALEEN WHALESFin 10000 0 555900 0 3500 0 194565 0Bowhead 0 100 0 3108 0 820 0 25482Gray 2300 0 35354 0 5000 0 76858 0Minke 3000 800 19698 5253 3000 800 19698 5253Right 100 0 2338 0 50 0 1169 0Blue 245 0 25170 0 160 0 16438 0Humpback 900 0 27367 0 1407 0 42784 0Sum 16545 900 665828 8360 13117 1620 351512 30735SPERM WHALES 12850 0 425399 0 6100 0 201941 0TOOTHED WHALESBeluga 8867 8867 2687 2687 8867 8867 2687 2687Killer 290 250 636 549 290 250 636 549Dall’s porpise 20000 10000 1226 613 20000 10000 1226 613Harbour porpoise 15000 7500 466 233 15000 7500 466 233Sum 44157 26617 5015 4081 44157 26617 5015 4081BEAKED WHALESStejnegers beaked 200 200 101 101 200 200 101 101Sum 200 200 101 101 200 200 101 101STELLER SEA LION 73000 73000 14279 14279 48000 48000 9389 9389WALRUS & BEARDEDPacific Walrus 1725 34500 1012 20234 3335 66700 1956 39120Bearded seal 5000 150000 1000 30000 5000 150000 1000 30000Sum 6725 184500 2012 50234 8335 216700 2956 69120SEALSNorthern fur 1561245 315081 38561 27312 690297 139284 17050 12076Harbour seal 33000 33000 2099 2099 18000 18000 1145 1145Spotted seal 157500 22500 7001 1000 157500 22500 7001 1000Ringed seal 0 400000 0 17000 0 400000 0 17000Ribbon seal 55000 55000 3930 3930 55000 55000 3930 3930Sum 1806745 825581 51590 51341 920797 634784 29125 351511950s 1980sNumbers Biomass (tons) Numbers Biomass (tons)76Mean Weight.  Estimates of mean body weight (wet, i.e., live weight) for malesand females of each species were obtained from Trites and Pauly (1998).   Thesewere applied to total population estimates to derive total biomass (Table A2.1).Ration.  Unless otherwise stated, individual ration (R, in percent of body weightper day) was estimated for each sex and species using:R = 0.1W0.8where W is the mean body weight in kg, 0.8 is from equation 23 in Innes et al.(1987),  and 0.1 is a downward adjusted value (from 0.123 in Innes et al. 1987),which accounts for the difference between ingestion for growth and ingestion formaintenance.Q/B ratios.  Annual consumption to biomass ratios were calculated for eachspecies based on their average body weight and the yearly ration. For the modelinput, a weighted average Q/B was calculated for each of the groups (TableA2.2). Q/B was weighted by biomass of each species to account for the largedifferences in the abundance of each species making up a group.Table A2.2.  The ratio of consumption (Q) to biomass (B) per year for marinemammals in the eastern Bering Sea during the 1980s.Group Q/BBaleen whales 11.38Sperm whale 4.55Toothed whales 13.10Beaked whales 10.51Walrus & Bearded 11.24Seals 15.95Steller Sea lion 12.70Diet.  Data on diet comes primarily from the 1996 NRC report (Tables 4.9 and4.10, p 129) and references therein (primarily Frost and Lowry 1981, and Lowryet al. 1982). Species consumed by marine mammals were matched to the 24aggregated groups identified above (see Pauly et al. 1998c).  Relative amountsconsumed of each group were determined from the weighted annualconsumption of the predators.  For example, 6 species of baleen whales belongto the baleen whale group.  As a group, they annually consume 2 million tonnesof pelagic and semi-demersal fishes, benthic invertebrates, large zooplankton andherbivorous zooplankton (Table: A2.3). We ranked the components of their dietas either major, minor or not eaten (2,1 and 0 respectively).  We then weightedthe importance of each component by the total amount consumed by eachspecies of whale and by all whales combined (i.e., weighted importance equalsthe sum of the product of consumption and importance divided by totalconsumption).  The proportion of the diet made up of the various prey typesequaled the weighted importance divided by the sum of all weighed importances.77Table A2.3.  Total consumption and the relative importance of prey types for 7 species of baleenwhales.  See text for details.Baleen Consump-Whales tion Pollock Benthic Epi- Cephal- Herb. Large Sum(t year-1) & pelag.p-feed. fauna opods zoop. zoop.Fin 1420325 2 0 0 1 2 2Bowhead 58750 0 1 0 0 2 2Gray 305975 1 2 2 1 0 0Minke 109451 2 0 0 1 1 2Right 4278 0 0 0 0 2 1Blue 119996 1 0 1 1 2 2Humpback 15837 2 0 0 0 0 2Sum 2034612Weighted imp. 1.73 0.33 0.36 0.96 1.63 1.70 6.71Proportion 0.26 0.05 0.05 0.14 0.24 0.25 1.00Importance1.  BALEEN WHALESGray Whale (Eschrichtius robustus)The eastern North Pacific population of Gray whales breeds along the west coastof North America and spends the summer feeding in the northern Bering, Chukchiand Beaufort Seas (Rice and Wolman 1971).  Some gray whales feed in watersof southeast Alaska, British Columbia, Oregon and Washington.  They are alsobelieved to feed in the eastern Bering Sea (D. Rugh, pers. comm.). In 1987-88the population was estimated at 20,869 (Small and DeMaster 1995) and rose to21,597 individuals in the mid 1990s (Hill et al. 1997)Fin Whale (Balaenoptera physalus)Fin whales, once the most abundant species of whales in the world oceans(Evans 1987), and commonly taken by whalers, are presently listed asendangered (Small and DeMaster 1995). Reliable estimates of abundance arenot available. Ohsumi and Wada (1974) report population ranges for the NorthPacific between 13,430 and 18,630 in the early 1970s following heavyexploitation.Over 9,000 fin whales were removed from the eastern Bering Sea from 1954 to1971 with most of the catch occurring in the eastern Aleutians near the easternBering Sea shelf break.  The maximum number killed in any one year was justunder 1,200 whales.Fin whale catch records look like those of other whale species (i.e., a high-sustained kills for a few short years, followed by a sharp collapse).  Given that finwhales have an estimated longevity of nearly 100 years (Ohsumi 1979), the totalharvest can be considered a minimum estimate of population size.  Assumingthat 10% of the population survived, suggests the eastern Bering Sea populationconsisted of 10,000 fin whales (9,000 harvested plus 1,000 surviving). Annualharvest was then calculated as 600 animals (9000 over 15 years).  This is78consistent with Ohsumi et al. (1971) who report 500 to 1000 taken annuallyduring the 1950s. Life table estimates suggest an equal sex ratio with meanweights of 59,819 kg for females and 51,361 kg for males (Trites and Pauly1998).  Note that these estimates assume that all age groups were present(calves through adults).The 1955 estimate of 10,000 fin whales is approximately 23% of the 42,000 to45,000 whales thought to be in the North Pacific.  Applying this estimate to the13,430 to 18,630 thought to be present in the 1970s suggest the Bering Seapopulation of fin whales was about 3,500 at that time.  This is higher than the1,000 fin whales estimated by Perez (1990) from relative abundance anddistribution data, but is consistent with the depletion estimate of abundance (i.e.,cumulative kill).Daily ration is about 40g per kg body weight per day during summer feedingseason (Lockyer 1981b). Approximate diet consists of 75% krill, 20% copepods,5% fish (Frost and Lowry 1981).Minke Whale (Balaenoptera acutrostrata)Minke whales have a wide distribution in both hemispheres and are found in theBering Sea mainly during spring and summer -- especially May and June --although some may stay all year (Ivashin and Votrogov 1981).  There are nopopulation estimates for the eastern North Pacific population (Small andDeMaster 1995).  Based on relative density data, Perez (1990) estimated a totalsummer population of  3,000 and a winter population of 800 minke whales in theeastern Bering Sea.  Diet consists mainly of pelagic and semi-demersal fish(pollock, herring, capelin - 60%), and euphausiids (30%) with some copepods(9%) and cephalopods (1%) – (Kasamatsu and Hata 1985).  No catches werereported for Minke whales in the region, so we assumed the same biomassvalues for both periods.Blue Whale (Balaenoptera musculus)According to Jefferson et al. (1993), the Gulf of Alaska is the northern limit forBlue whales. For this reason, 10% of those found in the North Pacific wereassumed to enter the southern part of the eastern Bering Sea. This gave apopulation of 160 for the 1980s.Blue whales were heavily depleted by the 1950s.  According to the NRC (1996)report, the initial population size of blue whales in the North Pacific was ca. 4,900animals. Whaling started in 1889, but presumably intensified by the 1940s and1950s with modern techniques.  In the Ecopath model, we considered that thepopulation size during the 1950s was half of the original population size, i.e., thetotal population size in the North Pacific was 2,450 animals. Only 10% of thisvalue was considered to be actively feeding in the Bering Sea, given their maindistribution is south of the Aleutian Islands (NRC 1996, Perez 1990). This gave apopulation of 245 for the 1950s. Braham (1991) reports a total of 5,761 bluewhales killed between 1889 and 1965, which gives an average catch of 75 peryear.  The average catch for the early 1960s was used, given an annual harvest79of 200 animals in the North Pacific.  This yields a catch of 20 whales per year  ifwe assume that 10% of these were taken in the eastern Bering Sea.Blue whales were assumed to consume 40g per kg of body weight per day duringsummer feeding season (Lockyer 1981b). Euphausiids are the major dietcomponent followed by copepods and nekto-benthonic invertebrates (Table 4.10in NRC 1996).Humpback  Whale (Megaptera novaeangliane)The size of the humpback whale population in the central North Pacific was 1,407in 1981 (Baker and Herman 1987) and 4,005 in 1996 (Calambokidis et al. 1997).The population is believed to have increased by as much as 10% per year fromthe 1980s to early 1990s (Hill et al. 1997), and may have numbered 15,000 priorto exploitation (Rice 1978). Commercial whaling is believed to have removed atotal 28,000 humpback whales.Perez (1990) estimated a Bering Sea population of 150 humpback whales in the1980s based on relative abundance and distribution data in Berzin and Rovnin(1966) and Wada (1980, 1981).  We assumed the 1950s eastern Bering Seapopulation was 6 times larger prior to exploitation than in the 1980s, i.e., 900whales (=6X150). This places 6% of the total North Pacific population in theeastern Bering Sea.  We used the average North Pacific catch (of 1,047 whales)in the early 1960s as an index of catches in the 1950s.  Again, we assumed that6% of these were taken in the eastern Bering Sea, given an annual harvest of 63whales.Based on summer feeding rates, humpbacks consume approximately 4% of theirbody weight per day (Lockyer 1981b). Diet consists mainly of euphausiids (69%)and fish (29% - Pacific herring, juvenile salmon, capelin, smelts, walleye pollock,Arctic cod, saffron cod, Pacific sand lance, rockfishes and Atka mackerel), as wellas small amounts of cephalopods (1%), copepods (1%), amphipods and otherinvertebrates (<1%) (Nemoto 1957, 1959, 1970; Kawamura 1980).Bowhead Whale (Balaena mysticetus)Bowheads were severely depleted during intense commercial whaling before the20th Century (Braham 1984).  The Western Arctic population was estimated at8,200 animals in 1993 (Zeh et al. 1994, 1995). The southern end of the Bowheadwhale winter distribution drops into the eastern Bering Sea, but it is a small part oftheir range.  We therefore assumed (based on Perez 1990) that 10% of thepopulation (820) occurred in mid-winter in the eastern Bering Sea in the 1980s.The 1996 population assessment reports the North Pacific population as currentlybeing 7,738 animals (Hill et al. 1997).Since Bowheads were severely depleted before the 20th century, it seems unlikelythat many were present in 1955.  We therefore assumed a 1955 population of100 individuals.80Bowheads feed primarily on euphausiids (45%), amphipods (27%) and copepods(24%) and occasionally ingest other invertebrates (4%) (Tomilin 1957, Johnson etal. 1966, Lowry and Frost 1984).Northern Right Whale (Eubalaena glacialis)The pre-exploitation size of the North Pacific Stock exceeded 11,000 animals(NMFS 1991).  Wada (1973) estimated a total population of 100-200 in the NorthPacific. We therefore assumed a population size of 100 in 1955. Since then,Soviet vessels killed approximately 5 right whales per year.  We thereforeassumed a 1980s population of 50 right whales.2.  SPERM WHALESThere are at least two populations of sperm whales (Physeter macrocephalus) inthe North Pacific - an eastern and a western Pacific population (Kasuya andMiyashita 1988).   Adult males from the western Pacific migrate north of thefemales following mating, unlike the eastern population, which does not appear tosegregate as strongly (Kasuya and Miyashita 1988). Between 40-60% of thesexually mature males from the western population are believed to migrate duringthe summer to high latitudes in the North Pacific, including the Bering Sea andAleutian Islands (Ohsumi 1966).  Time spent by whales on their summer feedinggrounds is 2-4 months based on information in Oshumi (1966).Male sperm whales become sexually mature when they reach 10-20 m in length(Lockyer 1976, 1981a; Gosho et al. 1984).  This corresponds to an estimated ageof 11+ years.  Applying weight at age curves calculated by Lockyer (1976) to thelife table estimates of Trites and Pauly (1998) indicates that adult males have amean weight of 33,105 kg.Gosho et al. (1984) indicate the eastern Pacific population of adult males (age11+ years) was 61,000 in 1982 and 128,500 in 1910. These animals were killedin small numbers from June to September, 1947-1954 (<200 per year) and atmuch higher rates from 1954-1966 (about 3,000 per year).  An almost knife-edgedrop in numbers killed after 1966 suggests that the western population wasseverely depleted by the late 1960s and no longer profitable to harvest.Distribution maps of kills suggest that the center of distribution of the males thatwent north was the central Aleutian Islands.  The maps further suggest that about10% were killed on the shelf edge and in the canyons of the eastern Bering Sea.This gives an annual harvest of 300 whales during the 1950s, and an estimatedeastern Bering Sea population of 6,100 adult male sperm whales in 1982 and12,850 in 1955.Perez (1990) estimated that the diet of sperm whales in the Bering sea consistsof 82% cephalopods (mostly squid) and 18% fish, with trace ingestion ofeuphausiids, shrimp, crabs, other invertebrates and marine mammals.  Fisheaten by sperm whales include salmon, lanternfishes, lancetfish, Pacific cod,walleye pollock, saffron cod, rockfishes, sablefish, Atka mackerel, sculpins,81lumpsuckers, lamprey, skates and rattails. These estimates were based on datain Berzin (1959), Okutani and Nemoto (1964), Tarasevich (1968) and Kawakami(1980),3.  TOOTHED WHALESBeluga Whales (Delphinapterus leucas)The two populations that occur within our study area are Bristol Bay and NortonSound (Small and DeMaster 1995). The 1996 population assessment reportclaims that most belugas in the Beaufort Sea, Norton Sound and Bristol Bay,overwinter and probably feed in the eastern Bering Sea and probably feed in thearea (Dave Rugh, NMML pers. comm.). The Bristol Bay population wasestimated at 1,500 from the mid 1980s to mid 1990s (Seaman et al. 1985, Frostand Lowry 1990, 1995).  These estimates are similar to those from the 1950s(Frost and Lowry 1990).    Norton Sound was estimated to have 7,367 belugas inthe early 1990s (Small and DeMaster 1995).  We have no reason to believe thatpopulations have changed much since 1955 and assumed a total population of8,867 belugas (1,500 + 7,367) in 1955 and 1980.  They are present 12 months ofthe year.  Winter populations may be augmented by animals from the easternChukchi (a population of 3,710).Belugas eat primarily pelagic and semi-demersal fishes (93%), but alsocephalopods (2%), euphausiids (3%), amphipods (1%) and other invertebrates(1%)  (Kleinenberg et al. 1964, Frost and Lowry 1981).  In Bristol Bay, salmonand smelts are eaten.  They may eat pollock in winter.Killer Whale (Orcinus orca)There are at least two forms of killer whales in the North Pacific.  One calledresidents eats fish, while the other called transients has specialized on eatingmarine mammals. Barrett-Lennard et al. (1995) estimated that male and femaletransients in the Gulf of Alaska consume 73kg of prey per day. This value wasassumed here for daily ration.Resident killer whales in the Eastern North Pacific population range from theChukchi and Bering Sea, along the Aleutian Islands, to the Gulf of Alaska andsouthward to California.  Photo-identification resulted in a minimum estimate of242 resident (fish eating) whales in the south-eastern Bering Sea, easternAleutian Islands and Kodiak region (Hill et al. 1997).  An additional 205 have beenidentified in Prince William Sound, 154 in Southeast Alaska, and 163 in BritishColumbia (for a total of 764 resident whales).  Population growth rates might beas low as 2%.  We assumed that the population has not changed much since1955 and set the resident population at 250 individuals.  Their diet is fish,including Pacific cod, skates, smelt, capelin, herring, halibut, sharks, salmon andArctic cod (Tomlin 1957, Sleptsov 1961).Transients.  Photo-identification indicates there are at least 36 transient (marinemammal eating) whales in the southeastern Bering Sea, eastern Aleutian Islandsand Kodiak region (Hill et al. 1997).  An additional 55 have been identified inPrince William Sound, 96 in Southeast Alaska, and 127 in British Columbia (for a82total of 314 transient whales).  Population growth rates are not known but areundoubtedly low. We assumed that the population has not changed much since1955 and set the transient population in the eastern Bering Sea at 40 individuals.Their diet consists of other marine mammals, primarily seals and sea lions butalso includes minke, humpback, gray, beluga whales; harbour and Dall’sporpoises (Tomlin 1957, Rice 1968, Ford et al. 1998).The total killer whale population of resident and transient killer whales in theeastern Bering Sea is 290 in both periods. The composite diet composition wasweighted according to the abundance of r sidents and transients.Dall’s porpoise (Phocoenoides dalli)The distribution of Dall’s porpoise extends to the Pribilof Islands, and may reachthe Bering Strait in summer (Perez 1990).  Limited surveys in Bristol Bay and thenorth Bering Sea resulted in an estimate of 9,000 (Hobbs and Lerczak 1993).Surveys north and south of the Aleutian chain gave an estimate of abundance of302,000.  Perez (1990) suggests the Dall population consists of 85,500 animals inthe summer and 42,700 in the winter.There is concern that boats may attract Dall’s porpoise and bias the estimates bya factor of five.  This means that these estimates should be multiplied by a factorof 0.2 if boats draw Dall’s porpoise to them and bias the estimate (Turnock andQuinn (1991).  We therefore assumed that the population did not changebetween 1955 and 1980 and that it consisted of  20,000 in summer and 10,000 inwinter.Demersal fishes, octopus and squids are thought to compromise the major part ofthe diet. Other components include pelagic and semi-demersal fishes,euphausiids and nekto-benthonic invertebrates (NRC 1996).Harbour porpoise (Phocoena phocoena)The distribution of harbour porpoise ranges from Point Barrow, along the Alaskancoast to southern California. In 1991, the Bristol Bay population was estimated at10,946 (Dalheim et al. 1992).  The survey did not include the Bering Sea andAleutian Islands.  This is significantly higher than the 1,000 estimated by Perez(1990) and does not include the Norton Sound population. The minimum numberfor the Bering Sea population from the 1996 population assessment report is8,549. Harbor porpoises that occur in the Chukchi Sea during summer probablywinter in the Bering Sea (Gaskin 1984).  Harbor porpoises are generally regardedas inhabitants of more inshore and shallower water compared to Dall’s porpoise.We have assumed a summer population of 15,000 and a winter population of7,500.  No change in numbers was assumed to occur from 1955 to 1980.Pelagic and semi-demersal fish form the major part of their diet. They also eatoctopus, squid and nekto-benthonic invertebrates.834.  BEAKED WHALESStejnegers beaked whale (Mesoplodon stejnegeri).This species is also known as the Bering Sea beaked whale (Loughlin and Perez1985) and is the most frequently encountered species of Mesoplodon (Rice1986).  It occurs in the deep waters of the south-west Bering Sea and canprobably be seen over the deep canyons that penetrate the Bering Sea shelf.There are no population estimates for this species.  Limited sightings suggest ayear round population of as few as 200 individuals (Perez 1990).  There is noinformation to suggest the population was any larger or smaller 30 years earlier.The primary food is probably squid (90%), but they may also feed on fish (10%)such as salmon.Baird’s beaked whale (Berardius bairdii).These whales have been sighted in the southern Bering Sea during summer(Tomilin 1957) in areas with submarine escarpments and seamounts (Kasuyaand Ohsumi 1984.  There is no information on abundance and harvest.5.  WALRUSES AND BEARDED SEALSPacific Walrus (Odobenus rosmarus divergens)The population was depleted as a result of intense commercial harvestingthrough the 1930s and 1940s.  The total population estimate (Russia and USA)was 65,500-94,400 in 1960 and 290,700-310,700 in 1980.  Annual harvests were4,500-6,500 in the 1950s, fell to 2,000-4,000 in the 1960s and 1970s, beforerising again to 6,000-9,000 in the 1980s.  The sex ratio of the harvest haschanged from 2.5 males: 1 female in the 1960s-70s to unity in the 1980s (Fay etal. 1997).The Bering Sea population was estimated to be 9,500 in 1960 and 66,700 in1980 (Fay et al. 1997).  Most of these animals spend the summer in the northernBering Sea and Bering Strait and move into the eastern Bering Sea in winter.Following Perez (1990) we assumed that 5% were present in summer and 100%in winter.Walrus prey largely upon benthic invertebrates, mostly infauna followed byepifauna and nekto benthonics. (Table 4.9 in NRC 1996).Bearded Seal (Erignathus barbatus)Bearded seals have a circumpolar range closely associated with sea ice. Theytend to use shallow areas where the ice is in constant motion. Estimates of theBering-Chukchi Sea population range from 250,000 to 300,000 (Popov 1976,Burns 1981a).  Perez (1990) estimated that 50% of the population occurs in theeastern Bering Sea, predominantly during November to April.  This suggests a1980 population of about 5,000 animals in summer and 150,000 in winter.  Weassumed the same number in 1955. Major prey items are similar to that of thewalrus but are very different from other seals. They prey largely upon benthicinvertebrates. Bearded seals are an important species for Alaskan subsistence84hunters, with estimated annual harvests of 1,784 animals from 1966 to 1977(Burns 1981a).6.  STELLER SEA LIONSSteller sea lions (Eumetopias jubatus) breed on offshore rocks and islands fromCalifornia to northern Japan. They general feed within 20 km of shore duringsummer. The size of the population was estimated by applying life table statisticsto counts of pups and adults made in the eastern Aleutian Islands (Trites andLarkin 1996).  Total population (including pups) was 73,000 animals in 1955 and48,000 in 1980.  Mean weights were 186 kg for females and 210 kg for males(Trites and Pauly 1998).  Sex ratios were 1.5:1.0 females to males.Stellers were assumed to be present 12 months of the year and eat a variety offish and cephalopods (Lowry et al. 1989, Merrick et al. 1997, Thorsteinson andLensink 1962, Calkins and Goodwin 1988, NRC 1996).  Walleye pollock are adominant food. Small pollock (<20cm) seem to be more commonly eaten byjuvenile sea lions than by older animals.  According to Merrick (1995), greaterthan 75% of the pollock eaten are juvenile fish. Other major diet componentsinclude, Atka mackerel, squid, herring, sandlance and rockfish.There is little information on the fluctuations of Steller sea lions that may haveoccurred before the 1960s. Harvest of Steller sea lions for the period wereconsidered to be 500 individuals annually (Trites and Larkin 1992).7.  SEALSNorthern fur seal (Callorhinus ursinus)Numbers of pups born on St. Paul and St. George Islands (Pribilof Islands) aretabulated in Lander (1980) and Trites (1989).  St. George is home to roughly 16%of the total Pribilof population (based on pup counts made between 1966 and1988).  This means that approximately 87,800 pups were born on St. George in1955 based on the 461,000 born on St. Paul.  Similarly the 203,825 pups countedon St. Paul in 1980 suggest a St. George population of 38,825.  Total size of thePribilof population (pups and non-pups) was 2.3 million in 1955 and 1.0 million in1980.  This is based on life tables calculated by Lander (1981), Trites and Larkin(1989) and Trites and Pauly (1998) indicating that pups make up 23.8% of thepopulation.  Sex ratios at birth are 50:50 (Trites 1991).  An unharvestedpopulation contains 38% males and 62% females (all ages combined).  Averageweight of all individuals (including pups) is 25.3 kg for females and 30.2 kg formales (Trites and Pauly 1998). During the commercial fur seal harvest, sex ratioswere approximately 35:65 males to females.85Table A2.4. Numbers of northern fur seals in the eastern Bering Sea during summerand winter.Year Season Males Females Total1955 Summer 331,1721,230,0691,561,2411955 Winter 315,018 0 315,0181980 Summer 146,427 543,871 690,2971980 Winter 139,284 0 139,284Pelagic distribution data (Bigg 1990) suggests that most immature animals ofages 1 and 2 years remain outside the Bering Sea during summer, and that about10% of the non-pups summer population remains during winter (Perez 1990).  Ofthe winter animals, most are adult males and some are immature animals.Females and pups receive their nutrition from the Bering Sea for approximately 6months (June through November).  Other than pups, males probably do not feedvery heavily during this time, having achieved most of their growth during thespring northward migration through the Gulf of Alaska (Trites and Bigg 1996).During summer (Table A2.4), we assumed that the female population consistedof pups and animals aged 3+ years.  This portion of the population makes up82% of the female population and has a mean weight of 28.22 kg.  We assumedthat only male pups and an average of 25% of males between the ages of 2 and7 years drew nutrition from the Bering Sea during summer.  This consists of 41%of the male population with an average weight of 11.62 kg.During winter (November to April), only mature males and some immature maleswere assumed to be present (10% age 3 years, 20% - 4, 35% - 5, 60% - 6, 80% -7, and 100% age 8+).  This represents 21% of the male population (which is 7.5%of the total population or 10% of the non-pup population) with an average weightof 86.7 kg.Dietary information was based on that provided in the NRC report (1996, theirTables 4.9 and 4.13). These were in turn based on the following references: Frostand Lowry 1981, and Lowry et al. 1982, Lucas 1899, Perez and Bigg 1986,Sinclair et al. 1994). Dominant food items in the Bering Sea are pollock squid,capelin and pelagic nektonic invertebrates. In the eastern Bering Sea, pollock,squid and capelin counts for 70% of diet  (Perez and Bigg 1981, 1986). There is apositive correlation between pollock year class strength and the amount of pollockin the diet. Pollock represent >80% of the diet of Northern fur seals, 96% of whichare juveniles (Sinclair et al .1994). Harvest of northern fur seals for the period wasconsidered to be 22,000 individuals per year (based on numbers killed reported inTrites 1989).Harbour Seal (Phoca vitulina)Harbour seals occur in coastal waters of the north Pacific from Baja Californiathrough the Aleutian Islands to Japan. In the early 1990s, the Bering Seapopulation, residing primarily in Bristol Bay and the north side of the AlaskaPeninsula, was estimated to be 18,322 animals (Small and DeMaster 1995).86Withrow and Loughlin (1996) estimated a 1995 population of 13,312 harbourseals in the Bering Sea.The overall Bering sea population is thought to be declining (Small and DeMaster1995) as it is in the Gulf of Alaska (Pitcher 1990).  Data are not available on thenumber of animals present in 1955 or to what extent the animals werecommercially hunted (estimates of annual kills range from 2,500 to 12,000 -- NRC1996).  We assumed the mid-range value (8,000) as the annual harvest for the1950s.An 83% decline of harbour seals has occurred at Sea Otter Island (PribilofIslands) since 1974.  Counts on the north side of the Alaska Peninsula in 1995were less than 42% of the 1975 counts.  The NRC (1996) panel felt that the1980s population was about 55% of the 1960s; hence, we assumed there were33,000 harbour seals in 1955 and 18,000 in 1980.  Mean weights, for all agescombined, were 58.4 kg for females and 68.6 kg for males (Trites and Pauly1998).Little is known about the food of harbour seals in the Bering Sea (Sease 1992).Available data are summarized in Table 4.15 of the 1996 NRC report (based onLowry et al. 1982). Major food items include herring, sand lance, smelt, sculpins,capelin, shrimp, mysids and octopus.Largha ‘Spotted’ Seal (Phoca larga)The primary range of the largha seal includes the Okhotsk, Bering and ChukchiSeas.  Burns (1973) estimated a world population of 335,000 - 450,000, with aBering Sea population of 200,00 - 250,000 (Lowry and Frost 1981, Lowry 1985,Burns 1986).  Perez (1990) suggests that 70% of the population occurs in theeastern Bering Sea from November to April based on relative abundance data bylocation presented by Braham et al. 1984.  He further suggests that 10% of thepopulation remains in this region from May to October.  This converts to 157,500animals (=0.7*225,000) during winter and 22,500 (=0.1*225,000) in the summer,and compares to an estimate from known August haul-outs of 59,214 animals(Small and DeMaster 1995).  Largha seals have never been commerciallyexploited.  Since it is not known whether this species has experienced changes inabundance similar to the better-studied species, we assumed no change since1955. Mean weights were 38.9 kg for females and 50.0 kg for males (Trites andPauly 1998).Largha seals feed primarily on fish (capelin, Pacific herring and walleye pollock --Lowry and Frost 1981, Lowry et al. 1982, Bukhtiyarov et al. 1984).  Some seals,especially young individuals also eat amphipods, shrimps, euphausiids, crabs,mysids and octopus.Ribbon Seal (Phoca fasciata)Burns (1981b) estimated the worldwide population at 240,000 in the mid-1970s,with an estimate for the Bering Sea at 90,000-100,000.  The animals can befound in the Bering Sea in winter, but are much further north in the summer.Perez (1990) estimates that 60% of the population are in the eastern Bering Sea87during November -April.  This suggests a population of about 55,000 in both timeperiods.  Major diet items are fishes, both pelagic and demersal (Table 4.9 inNRC 1996).Ringed Seal (Phoca hispida)World population estimates range from 2.3 to 7.0 million, with 1.0 to 1.5 million inAlaskan waters (Kelly 1988).  The winter distribution of ringed seals suggestsabout one-third of the population (400,000) moves into our study area duringwinter.  This compares to the feelings of Lowry et al. (1982) and Frost (1985) thatat least 250,000 ringed seals occur on the shorefast ice in the Bering Sea. Theyare thought to have a similar diet to spotted seal; i.e., they eat semi-pelagic fishand pelagic and benthic invertebrates.8.  PISCIVOROUS BIRDSThe main species of piscivorous birds in the eastern Bering Sea are the northernfulmar, black-legged and red-legged kittiwakes, and common and thick-billedmurres.  Population trends at the main nesting sites of these birds on the PribilofIslands (St. George and St. Paul) are available from counts performed during fouryears considered by the model between 1979 and 1985 (Climo 1993, Dragoo andSundseth 1993).  Population estimates were transformed into biomass using themean individual weights of these species found in Hunt et al. (1981).  The valuefor daily ration used for these species was 20% of their body weight per day (Huntet al. 1981), which is probably conservative (Schneider and Hunt 1982).  Weassumed these species of birds fed in the eastern Bering Sea for approximately300 days per year and that they consumed primarily juvenile pollock,euphausiids, and other fish (Hunt et al. 1981).  Based on survival estimates ofadult black-legged kittiwakes reported by Hatch et al. (1993), adult mortality ratesare approximately 0.4 year -1 and total population mortality rates are most likelyhigher.There is no information available on the abundance of marine birds during the1950s.  According to George Hunt (pers. comm) there is no basis for claiming amajor decrease in seabirds abundance before 1976.  Dramatic die-offs of marinebirds in the eastern Bering Sea are reported for the period between 1976 and1984.  The decline in  abundance of murres in the Pribilofs Islands were in theorder of 25% between mid-1970s and mid-1980s (A. Springer unpubl. data).Given the above, marine birds biomass was considered the same in the 1950sand 1980s models.Fish and CephalopodsThe eastern Bering Sea supports large populations of groundfish, includingwalleye pollock, Pacific cod, small and large-mouthed flounders, skates, andsculpins.   Over 90% of the fish  biomass estimated from bottom trawl survey ofthe shelf area in 1985 was composed of walleye pollock, yellowfin sole, Pacificcod, rock sole, Alaska plaice, flathead sole, sculpins, skates, and arrowtooth88flounder (Walters et al. 1988, Williamson and Smith 1988).  Dominant groundfishpopulations in the slope regions include giant grenadier, walleye pollock,Greenland turbot, arrowtooth flounder, sablefish, Pacific ocean perch and Pacificcod.  Stock assessments are regularly performed for the major groundfishspecies in this region to provide advice on acceptable biological catch limits to theNorth Pacific Fishery Management Council (NPFMC 1996).  These populationassessments are the primary source of catch, biomass, and mortality estimatesfor the 1980s model.Food habits data on the main species of groundfish in the eastern Bering Seahave been collected by the Alaska Fisheries Science Center since about 1981and on a regular basis since 1984.  There are several reports that summarize thedata used to provide diet information for groundfish for the 1980s model (seeLivingston and Goiney 1983, Livingston et al. 1993, Dwyer et al. 1987, Brodeurand Livingston 1988, Livingston 1991).  These data primarily reflect diet com-position during the main feeding season in summer.  The food habits data base atthe Alaska Fisheries Science Center was used to estimate the diet of yellowfinsole, flathead sole, Pacific cod, sablefish, walleye pollock, arrowtooth flounder,Greenland turbot, rock sole, Alaska plaice, eelpouts, Pacific halibut, rockfish,Pacific herring, sculpins, and skates.For the 1980s models, population diet composition was calculated by weightingthe diet of predator size groups in geographic strata by the biomass of those sizegroups by area.   Overall diet was then calculated using the biomass of each sizegroup as a weighting factor.  Estimates of daily rations followed the approachdescribed in Livingston (1991), which uses information on annual growthincrements and conversion efficiencies to derive annual rations for a given agegroup.  For fish populations with a known age composition, rations were derivedfor each age group and an overall ration was estimated from the average agecomposition of the population during the 1979 to 1985 period.For the 1950s model, two sources of information were used to calculate thebiomass and harvest rates of finfish (Table A2.5). Reports of Soviet surveys in theBering Sea were used to calculate species biomass (Vidar Wespestad, NMFSpers. comm.).  Catches were obtained from Bakkala (1993).Table A2.5. Finfish catches and biomass in the 1950s.Species Group       Biomasst km-2          proportion                   of 1980s          Catcht km-2         proportion                  of 1980sAdult pollock 4.320 0.157 0.014 0.007Juvenile Pollock 0.942 0.157    -    -Deepwater fish 1.011 2.484 0.0001 0.013Large flatfish 1.169 0.615 0.002 0.019Small flatfish 8.593 0.937 0.105 0.322Other demersal fishes1.146 0.294 0.001 0.007Pelagics    -    - 0.0826 0.39089Particular aspects of each species' biology and parameter estimation details arenoted below.9,10.   POLLOCKWalleye pollock (Theragra chalcogramma) were broken into two groups: juvenile(ages 0 and 1) and adults (ages 2 and older) because of the cannibalism thatoccurs primarily on age 0 and age 1 fish.  Estimates of juvenile biomass andmortality were derived from age-1 pollock biomass estimated from an integratedcatch-at-age population model of pollock which included cannibalism andpredation by Pacific cod and northern fur seals (Livingston and Methot 1998).Age-0 biomass was derived by back-calculating estimated age-1 numbersassuming half-yearly mortality rates for age-0 fish of 1.0, a conservative estimatecompared with estimates of 2.6-3.2 obtained by Livingston (1993).  Estimates ofadult biomass and mortality rates were obtained from the pollock populationassessment (Wespestad et al. 1996).Biomass of adult and juvenile pollock for the 1950s (Table A2.5) was calculatedas a proportion (0.157) of 1980s biomass based on surveys from 1957-60 (VidarWespestad pers. comm.). Adult catches were calculated as 0.014 t•km -2 basedon Bakkala (1993)11.  DEEPWATER FISHBiomass and mortality estimates for sablefish (Anoplopoma fimbria), rockfish(Sebastes spp.), and grenadier (Macrouridae) were obtained from the respectivepopulation assessment reports for each group (Fujioka et al. 1996, Ito and Ianelli1996, Ito 1996, and Fritz 1996).  Sablefish rations were estimated as describedabove.  Rockfish and grenadier rations were assumed to equal the sablefishrations.  Diet for sablefish and rockfish was estimated from the food habits database at the Alaska Fisheries Science Center as described above.  Grenadier dietwas taken from Novikov (1970).Biomass for the 1950s was calculated as 1.011 t•km-2, from Bakkala (1993) andcatches were 49.66 t , i.e., 0.0001 t•km-2 (Table A2.5).12.  LARGE FLATFISHPacific halibut (Hippoglossus stenolepis) population size and mortality parametersare estimated by the International Pacific Halibut Commission (IPHC) andreported in the North Pacific Fishery Management Council's populationassessment document (IPHC 1996).  Greenland turbot (Reinhardtiushippoglossoides) and arrowtooth flounder (Atheresthes stomias) populationbiomass and mortality were obtained from Ianelli et al. (1996) and Wilderbuer andSample (1996).  Because food habits sampling for estimating Pacific halibut diet90did not begin until 1989, diet information for the 1980s model was taken from dataobtained from 1989 through 1992 in the eastern Bering Sea.Biomass of large flatfish (1.169 t•km-2) for the 1950s was calculated as aproportion (0.615) of 1980s biomass based on surveys from 1957-60 (VidarWespestad). Catches (820.5 t, i.e., 0.0016 t•km-2) were taken from Bakkala(1993) (see Table A2.5).13.  SMALL FLATFISHSmall-mouthed flounders, which feed primarily on benthic infauna and epifauna,are a major presence in the inner and middle-shelf regions of the eastern BeringSea.  Biomass and mortality population parameters for yellowfin sole(Pleuronectes asper), rock sole (Pleuronectes bilineatus), Alaska plaice(Pleuronectes quadrituberculatus) and flathead sole (Hippoglossoides elassodon)were obtained from Wilderbuer (1996), Wilderbuer and Walters (1996), andWalters and Wilderbuer (1996).Biomass of small flatfish  (8.593 t•km-2) for the 1950s was calculated as aproportion (0.937) of 1980s biomass based on surveys from 1957-60 (VidarWespestad, pers. comm.). Catches (50,928 t = 0.105 t•km-2) were taken fromBakkala (1993) (see Table A2.5).14.  PELAGICSThe Pelagics group consists primarily of Pacific herring (Clupea pallasi), capelin(Mallotus villosus), sandlance (Ammodytes hexapterus), lanternfish(Myctophidae) and bathylagids (Bathylagidae).  Biomass estimates exist forPacific herring (Wespestad 1991).  However, the remaining species are not well-sampled in the bottom or mid-water surveys of the eastern Bering Sea due eitherto the mesh size of the nets used or to the lack of attention given to inshoreareas.  Total mortality rates were initially assumed to be those published forherring (Wespestad 1991), which range from about 0.3-1.2 year-1 depending onage.  A value of 1.0 year-1 was used for herring, which is commercially fished, and0.8 year -1 for the other pelagics, which are not a target for fisheries.  Ration anddiet composition were calculated for Pacific herring as described above and wasused to describe the group as a whole.  Because biomass is poorly known for thisgroup, ecotrophic efficiency was fixed at 0.9 so that biomass could be estimatedby Ecopath.Herring biomass during the late 1950s was between 600,000 and 900,000tonnes.  The high biomass values were mainly due to two large year-classesduring 1957 and 1958 (Wespestad 1991).  Biomass during the mid-1970s was inthe order of 100,000 tonnes.  There is no information available on biomass ofsandlance, capelin and myctophids.  Salmon biomass during the 1950s was ca.57,000 tonnes, and 147,000 tonnes for the 1975-1980s period (NRC 1996, p.111,Fig. 4.28). Pelagics biomass was not specified in the models due to the lack ofinformation on some pelagic species.  Nevertheless, the information available91allowed us to estimate a minimum biomass value for the group in both periods:757,000 tonnes (1.562 t•km-2) for the 1950s, and 650,000 tonnes (1.342 t•km-2)for the 1980s.Catches of herring intensified in the Bering Sea during the 1960s and peaked inthe early 1970s, with a brief but intense foreign fleet winter trawl fishery on theoffshore winter grounds (NRC 1996). Catches were in the order of 10,000 t in1960 (Wespestad 1991). Catches of salmon species in the eastern Bering Sea inthe period of 1975 to 1980 varied between 10 and 40 million fish (NRC 1996).During the 1950s catches were between 5 and 15 million fish.  To convertcatches in numbers to weight we used a value of 3 kg as the mean weight ofsockeye salmon (the most important species) in the North Pacific Ocean andBering Sea (Groot et al. 1995).  Catches in weight were then 75,000 t in the1980s and 30,000 tonnes in the 1950s.  The annual harvest of Pelagics in the1950s was estimated to be 40,000 t (0.0826 t•km-2).15.  OTHER DEMERSAL FISHSpecies in this group include Pacific cod (Gadus macrocephalus), skates(Rajidae), sculpins (Cottidae), and eelpouts (Zoarcidae). Catch, biomass andmortality estimates for cod were obtained from Thompson and Dorn (1996).Catch and biomass of skates and sculpins are reported in Fritz (1996) andeelpout biomass was obtained from Gary Walters (personal communication,Alaska Fisheries Science Center, 7600 Sand Point Way NE, Seattle, WA 98115).Mortality estimates were initially assumed equal those of cod (0.4 year-1 ), butwere later adjusted upward to 0.6 year-1 for eelpouts because of the largepredation on them by cod.  Rations were estimated for cod and skates asdescribed above.  Rations of skates and eelpouts were assumed to be similar toskates.Biomass in the 1950s was taken as the sum of cottidae and cod values from theSoviet surveys [cottidae =  0.790 t•km-2 (0.532 of 1980s biomass); cod = 0.356t•km-2  (0.125 of the 1980s biomass); total 1.146 t•km-2]. Catches (505.83 t, i.e.,0.001 t•km-2) were taken from Bakkala (1993) (see Table A2.5).16.  CEPHALOPODSAlthough there are several species of cephalopods in the eastern Bering Sea, weassumed the dominant species is Berryteuthis magister.  Catch estimates ofsquid in the eastern Bering Sea are from Fritz (1996).  Biomass, mortality andconsumption rates and diet were obtained from Radchenko (1992).  Becausedaily ration estimates for squid reported by Radchenko (1992) encompass a largerange  (1.1-4.2% body weight daily), gross conversion efficiency for squid was setat 0.3 so that ration could be estimated by the model. Biomass was consideredthe same for the 1950s model since no information could be found.92Benthics and Jellies17.  BENTHIC PARTICULATE FEEDERSBenthic particulate feeders include snow and Tanner crabs (Chionoecetes opilioand C. bairdi), red and blue king crabs (Paralithodes camtschatica and P.platypus), and shrimp (particularly Pandalidae and Crangonidae).  Biomassestimates for snow, Tanner, and king crabs were obtained from the summerbottom trawl survey of the eastern Bering Sea performed by the Alaska FisheriesScience Center (see summaries of these data in Otto et al. 1997).  Shellfishcatch data for the 1980s model were obtained from ADF&G (1994) and catchesfor the 1950s model from Otto (1986).  Rations for crabs were obtained using thesize-specific rations of Tanner crab in Paul and Fuji (1989) and weighting theoverall ration by the size composition of each crab population.  Ration for shrimpwere based on estimates presented by Evans (1984) for Crangon crangon, acongener of the crangonid species found in the eastern Bering Sea.  Dietcomposition for crab was obtained from Pearson et al. (1984) with increases indetrital contribution to the diet based on information presented in Brethes et al.(1994).  Shrimp diet was estimated from information in Feder (1978) and Rice(1981).  Mortality rates for crab were obtained from a variety of sources: JerryReeves (personal communication, Alaska Fisheries Science Center, 7600 SandPoint Way NE, Seattle, WA 98115), Fukuhara (1985), Somerton (1981) andLivingston et al. (1993).Biomass of king crab during the 1950s was approximately 5 times larger than inthe 1980s (NRC 1996, p. 75). This was assumed true for other species of craband shrimps. Final biomass value in Ecopath is 29 t•km-2.  Harvesting effort andcatches declined in the 1950s followed by a period of low and variable catchesthrough to 1966, before expansion to the current, full-scale fishery.  Catches in1966 were about 1 million lb. (ca. 460 t). Catches of particulate feeders for the1950s were assumed to be 0.0098 t•km-2, based on figures from 1953-1959reported by Otto (1986).18.  INFAUNAInfauna consists of clams, polychaetes, and other worms (mainly Echiuridae).Surveys of infaunal biomass in the eastern Bering Sea have rarely been done.There are two primary sources of information on infaunal biomass: a survey doneby Soviet scientists in 1958 (Neiman 1968) and a survey performed for the OuterContinental Shelf Environmental Assessment Program (OCSEAP) (Haflinger1981, McDonald et al. 1981) in 1975-1977.  Although the OCSEAP programsamples are more recent, they encompassed a smaller portion of the easternBering Sea shelf than the Soviet samples.  In addition, the OCSEAP sampleswere taken at the end of what might be considered a cold period in the easternBering Sea while the Soviet samples were from more of an average period(Niebauer 1988) that might be more representative of the 1980s model period.Initial biomass values for these groups for the 1980s model were taken fromNeiman (1968).  The diet of these infaunal groups were assumed to be 100%detritus based on the food web information presented by Feder and Jewett93(1981).  P/B estimates were not available for Bering Sea infaunal groups while awide range of estimates were available for infauna from other regions (Banse andMosher 1980, Evans 1984, Seitz and Schaffner 1995, Sissenwine et al. 1984,Warwick 1980).   The estimate  of  P/B= 1.3 year-1 for clams was obtained fromEvans (1984) and initial P/B=3.0 year-1 for polychaetes and worms from Seitz andSchaffner (1995) but was later changed to the 1.5 year-1 value reported bySissenwine et al. (1984).  Similarly, rations for these groups have not beenestimated for the Bering Sea and the value used (Q/B=12 year-1) was derivedfrom Evans' (1984) estimate for clams.19.  JELLYFISHBased on observations of taxonomic composition of trawl catches in recent years,the majority of scyphozoa in the eastern Bering Sea are believed to beChrysaeora melanaster. Jellyfish biomass (0.048 t•km-2) was derived fromaveraging NMFS trawl survey catches of medusae from 1982 to 1985.  Jellyfishdiet is assumed to consist of: 13.9% large zooplankton, 71.5% small zooplankton,0.8% jellyfish, 1% juvenile pollock, and 12.8% crab larvae (Hamner 1983).  Rationestimate of 1.09 g per gram body weight per year was derived from information inArai (1997) on ration of Chrysaeora sp. in terms of nitrogen and converted to wetweight using data on water and nitrogen content of jellyfish and jellyfish prey.  P/B(0.857 year-1) was assumed equal to the inverse of the generation time (Allen1971) of 14 months (Arai 1997).  The biomass was assumed to be present in the1950s. Q/B was subsequently increased to 2.0 year-1 while balancing the model.20.  EPIFAUNASpecies included in epifauna are hermit crabs, snails, brittlestars, and starfish.Biomass estimates for hermit crabs, snails, and starfish were obtained from thebottom trawl survey estimates of these groups for the eastern Bering Sea shelf,averaged from 1979 to 1985.  Brittlestar biomass estimates were from Neiman(1968).  Diet compositions were obtained from the OCSEAP studies ofinvertebrates summarized by Feder and Jewett (1981).  P/B estimates were notavailable for Bering Sea species of infauna so the relationship between annualP/B and life-span derived for marine benthic invertebrates in Warwick (1980) wasused, assuming most organisms in this group have life-spans of between 2 to 5years.  Ration estimates were not available for these groups, so estimates closeto those of crabs were used.No information was available on for the 1950, hence it was assumed the same as1980s.94Plankton21.  LARGE ZOOPLANKTONThis group includes euphausiids, mysids, and amphipods.  Initial biomassestimates for euphausiids of 10 t•km-2 were obtained from English (1979).  Abenthic amphipod biomass estimate (2.2 t•km-2) was available from Neiman(1968) and is undoubtedly an underestimate of the total amphipod (pelagic andbenthic) biomass.  No initial mysid biomass estimate was available so it wasinitially assumed to be similar to the amphipod biomass.  A P/B estimate for eachof these animals based on body size considerations (Banse and Mosher 1980)would be around 2.5 - 4.0 year-1 for an animal with a mean body weight (inenergetic equivalents) of 0.02 kcal.  The production and biomass estimatespresented in Cooney (1981) for the Bering Sea zooplankton communities in theinner, middle, and outer shelf regions were converted into P/B ratios andcombined into a total P/B=6.46 year-1 for the whole eastern Bering Sea shelf andslope region.  Initial estimates of P/B=2.5 year-1 were used for each group andthen later increased.  Initial rations for large zooplankton were derived fromestimates for Bering Sea copepods of 33.2 g per gram body weight per year(Dagg et al. 1982).  These rations were later decreased to 22.0 g/g/year-1 toaccount for possible seasonal declines in or cessation of feeding.  Diets of mysidsand euphausiids were obtained from Mauchline (1980) and diet of amphipodsfrom Barnes (1980).Meshcheryakova (1970), estimates total zooplankton biomass (from samplestaken 1961-1965) to be 42 t•km-2.  In the 1980s model we have split zooplanktoninto large zooplankton and herbivorous zooplankton (copepods).  Both groupshave biomass close to those reported by Meshcheryakova (44 and 55 t•km-2,respectively).  There is some uncertainty about what Meshcheryakova (1964,1970) considered as total zooplankton biomass.  If it refers to small zooplankton,then we have a good indication that zooplankton biomass probably remained thesame between the 1950s and 1980s. On the other hand, data on primary andsecondary production derived from baleen isotope analyses (see 23.Phytoplankton) point to higher values for the 1950s.  We assumed the biomassand production of large and herbivorous zooplankton did not differ between thetwo modeled time periods.22.  HERBIVOROUS ZOOPLANKTONThis group consists primarily of small and large calanoid copepods of the easternBering Sea slope and shelf regions.  Initial biomass estimates of 46.9 kg•km-2were obtained from Motoda and Minoda (1974) and later increased to 55kg•km-2.  As with large zooplankton, estimates of P/B range from 2.50-6.46 year-1(Banse and Mosher 1980, Cooney 1981). A P/B estimate in the upper end of thisrange (6 year-1) was chosen for this group to reflect its smaller size and fasterturnover rate relative to large zooplankton.  Rations for copepods were the sameas those for large zooplankton.  The diet was assumed to be composed of bothphytoplankton and detritus.9523.  PHYTOPLANKTONEstimates of phytoplankton production (primary productivity) for the easternBering Sea range from around 160 g C m-2 year-1 for the middle and outer shelfareas to lower values of around 75 g C m-2 year-1 for the inner shelf region(Springer et al. 1996).  Values range widely (175-890 g C m-2 year-1) for the areaof high production around the shelf edge region (Springer et al. 1996).  Sincemany of the estimates are acknowledged to be minimum estimates (e.g., Walshand McRoy 1986), we initially used values from the upper end of the range (i.e.,236.5 g C m-2 year-1 from Iverson and Goering (1979), converted to 1337 t wwkm-2 year-1, assuming 0.4 g C per g dry weight and 0.5 g dry wt per g wet weight).This value was adjusted to 1264 t ww km-2 y-1 to reflect our delineation of theeastern Bering Sea (Fig. 1).Walsh and McRoy (1986) estimate that 17% of the annual primary production onthe middle shelf is buried (i.e., lost).  They estimate that as much as 49% in theouter domain is exported to the slope where it is either buried or consumed. It isnot clear what proportion of this is available to consumers and what portion isburied or exported beyond our model boundaries.Primary production that is buried or exported to other areas within the system isassumed to accumulate as detritus, and is thus not lost from the system. Toaccount for actual transport or export of unutilized primary production (as notedby Springer et al. 1996) we specified that 10% of the primary production that endsup in detritus is lost (exported) from the system. This was implemented in thedetritus fate (0.9).  Changes in primary production in the Bering Sea areuncertain.  Baleen isotope data has shown that there has been a significantdecline in primary and secondary production (Don Schell, University of Alaska,pers. comm.).  The decline has been in the order of 35 to 40% from 1950s to1980s.  We therefore used two different values for primary production: 1264 t wwkm-2 year-1 for the 1980s and 1770 t ww km-2 year-1 for the 1950s. The initialestimate of turnover rate P/B=170 year-1 w s derived from the daily turnover ratesof diatoms reported in Motoda and Minoda (1974) and expanded to a half-yearlyproduction period.  However, specific values for biomass and P/B ratio forphytoplankton were adjusted during the balancing of the models (see below).Other24.  DISCARDS AND BY-CATCHThe discard box was created to explicitly account for by-catch and discards.Catches reported from the Bering Sea already include discards and weredisentangled using discard rates (the estimated percentage of total catch that wasdiscarded) from 1990-1994 (Queirolo et al. 1995, their Table 55). Catches anddiscards were further explicitly assigned to different fishing gears using theproportions of the target and by-catch species catches retained and discarded byfisheries in the Bering Sea/Aleutian Islands during 1993 (Alverson and Hughes961995). Gears were divided in the 5 types: Bottom Trawl (including the fisheriesfor pollock, Pacific cod, rock sole, Pacific Ocean perch and yellowfin sole); MidWater Trawl (fishery for pollock); Long Line (including the fisheries for sablefish,Pacific cod and halibut); and Pots (including fisheries for king crab, Bairdi tannerand Opilio tanner).  Shooting and entanglement of marine mammals wereassigned as a sixth type of fishing.  Table A2.6 shows the fractions of the catchweight that were landed and discarded by gear type in the eastern Bering Sea.For each gear type, the catch of non-target species (by-catch) is input to thediscard box.Discards are directly incorporated as small proportions in the diet of some groups,particularly birds and pinnipeds. Although it is impossible to determine fromstomach content analysis of groundfish species whether a particular item wasconsumed while it was dead or alive, the following groundfish species have beenobserved consuming fish processing offal (fish remains thrown overboard):Pacific cod, skates, sablefish, walleye pollock, arrowtooth flounder, flathead sole,yellowfin sole and Pacific halibut (Queirolo, et al. 1995).  Of these species, Pacificcod, skates, sablefish and Pacific halibut would be the species most likely toconsume whole discarded fish.Table A2.6: Catches (t• km-2  year-1) retained and discarded by fisheries in the Bering Seaduring the 1980s (from Queirolo et al. 1995 and Alverson and Hughes,1995).Species Landed Discard Landed Discard Landed Discard Landed Discard Landed Discard Landed DiscardPollock 0.1288 0.1152 1.7658 0.0697 0.0004 0.0031 0.0000 0.0000 0.0000 0.0000 1.8950 0.1880Cod 0.0649 0.0156 0.0016 0.0043 0.0620 0.0028 0.0000 0.0000 0.0000 0.0000 0.1285 0.0227Halibut 0.0000 0.0015 0.0000 0.0003 0.0009 0.0003 0.0000 0.0000 0.0000 0.0000 0.0009 0.0021Grenland turbot 0.0307 0.0179 0.0020 0.0029 0.0143 0.0092 0.0000 0.0000 0.0000 0.0000 0.0469 0.0300Arrowtooth fl. 0.0011 0.0115 0.0001 0.0019 0.0005 0.0059 0.0000 0.0000 0.0000 0.0000 0.0017 0.0193Flathead sole 0.0027 0.0061 0.0001 0.0010 0.0000 0.0001 0.0000 0.0000 0.0000 0.0000 0.0028 0.0072Yellow sole 0.1882 0.0681 0.0001 0.0013 0.0001 0.0003 0.0000 0.0000 0.0000 0.0000 0.1883 0.0696Rock sole 0.0121 0.0180 0.0000 0.0009 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0121 0.0189Alaska plaice 0.0074 0.0165 0.0002 0.0028 0.0000 0.0002 0.0000 0.0000 0.0000 0.0000 0.0076 0.0195Herring 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0550 0.0000 0.0550 0.0000Other pelagics 0.0026 0.0008 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.1542 0.0000 0.1568 0.0008Sablefish 0.0001 0.0000 0.0000 0.0000 0.0047 0.0001 0.0000 0.0000 0.0000 0.0000 0.0049 0.0002Rockfish 0.0006 0.0002 0.0000 0.0000 0.0015 0.0007 0.0000 0.0000 0.0000 0.0000 0.0021 0.0009C. bairdi 0.0000 0.0018 0.0000 0.0002 0.0000 0.0000 0.0190 0.0139 0.0000 0.0000 0.0190 0.0159C.Opolio 0.0000 0.0044 0.0000 0.0006 0.0000 0.0000 0.0480 0.0352 0.0000 0.0000 0.0480 0.0402King crab 0.0000 0.0034 0.0000 0.0000 0.0000 0.0000 0.0410 0.0280 0.0000 0.0000 0.0410 0.0314Pelagic TotalBottom Trawl Midwater Trawl Long line Crab pots25.  DETRITUSThe detritus fate describes what happens to the detritus produced by each group(from unassimilated food and from decomposition of organisms). Under a steadystate assumption the ecotrophic efficiency of the detritus box should equal 1.0.Adjustments were made to the fate of detritus for each grouped based on thefollowing assumptions:· For marine mammals, only 25% of marine mammal detritus remains withinthe system.   This seems to be a reasonable assumption given that mostmarine mammals are only present in the eastern Bering Sea during thesummer;97· Birds were also set to 25% given that the corpses of most birds probably endup on land and not sea;· All of the detritus was considered to accumulate from those found in closeassociation with the bottom (infauna, epifauna, particulate feeders) or thoserestricted to the shelf area (small flatfish);· For all other groups, we assumed that 25% of detritus was exported from thesystem due to transport out of the shelf/slope area of the eastern Bering Seaand into the deep ocean (leaving 75% to accumulate);· Accumulated detritus was allowed to remain within the system (by assigning avalue of 1 to the fate of detritus for the Ecopath detritus box).98Appendix 3 – Parameters for the 45-Box Ecopath ModelTable A3.1. Ecopath parameters describing the 1980s eastern Bering Seaecosystem with 45 functional groups.Functional Biomass P/B Q/B Catch Group (t km-2) (year-1) (year-1) (t km-2)Baleen whales 0.39 0.02 11.38 0.0000Toothed whales 0.01 0.02 13.11 0.0000Sperm whales 0.21 0.02 4.55 0.0000Beaked whales 0.00 0.02 10.52 0.0000Walrus & Bearded 0.07 0.06 11.25 0.0091Seals 0.07 0.06 15.93 0.0013Steller sea lions 0.02 0.06 12.70 0.0001Pisc. Birds 0.01 0.80 60.00 0.0000Adult pollock 27.45 0.50 2.64 1.8950Juv Pollock 6.00 2.50 8.33 0.0000Cod 2.42 0.40 2.04 0.1285Halibut 0.14 0.40 2.49 0.0009Grenland turbot 0.96 0.40 2.04 0.0470Arrowtooth flounder 0.80 0.40 2.92 0.0017Flathead sole 0.43 0.40 2.56 0.0028Yellow sole 6.11 0.40 2.96 0.1884Rock sole 1.34 0.40 3.60 0.0121Alaska plaice 1.29 0.40 2.49 0.0076Herring 0.78 1.00 3.65 0.0550Others pelagics 13.22 0.80 3.65 0.1568Skates 0.29 0.40 2.56 0.0000Sculpins 0.56 0.40 2.56 0.0000Sablefish 0.11 0.40 2.49 0.0048Rockfish 0.09 0.40 2.49 0.0021Macrouridae 0.20 0.40 2.49 0.0000Zoarcids 0.64 0.60 2.49 0.0000Cephalopods 3.50 3.20 10.67 0.0000C. bairdi 0.60 1.00 5.00 0.0190C.opilio 1.60 1.00 5.00 0.0480King crab 0.60 0.60 5.00 0.0410Shrimp 3.00 2.04 10.20 0.0000Clams 29.50 1.30 12.00 0.0000Polychaetes 14.00 1.50 12.00 0.0000Other worms 3.00 1.50 12.00 0.0000Hermit crab 1.00 1.80 8.00 0.0000Snail 0.52 1.80 8.00 0.0000Brittlestar 3.00 1.50 5.00 0.0000Starfish 1.34 1.50 5.00 0.0000Amphipods 6.00 3.50 22.00 0.0000Jellyfish 0.05 0.88 2.00 0.0000Euphausiids 35.00 5.50 22.00 0.0000Copepods 55.00 6.00 22.00 0.0000Mysiids 3.00 3.50 22.00 0.0000Phytoplankton 32.00 60.00 0.00 0.0000Discards 0.00 - - 0.0000Detritus 0.00 - - 0.000099Appendix 4 – Diet Matrix TablesTable A4.1.  The estimated proportion of prey (nos. 1-24) eaten by predators (nos. 1-11) during the 1950s.Prey1 2 3 4 5 6 7 8 9 10 11  1. Baleen whales 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  2. Toothed whales 0.000 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  3. Sperm whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  4. Beaked whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  5. Walrus&Bearded 0.000 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  6. Seals 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  7. Steller Sealion 0.000 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  8. Pisc. birds 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  9. Adult pollock2+ 0.002 0.008 0.002 0.002 0.002 0.002 0.002 0.000 0.000 0.000 0.020 10. Juv. pollock0-1 0.007 0.013 0.002 0.006 0.002 0.018 0.029 0.095 0.020 0.000 0.008 11. O.Dem.Fish 0.000 0.056 0.050 0.000 0.016 0.052 0.031 0.106 0.007 0.000 0.043 12. Large Flatfish 0.000 0.053 0.000 0.000 0.013 0.046 0.003 0.000 0.001 0.000 0.000 13. Small Flatfish 0.000 0.050 0.000 0.000 0.010 0.040 0.003 0.000 0.000 0.000 0.120 14. Pelagics 0.182 0.319 0.056 0.172 0.044 0.288 0.862 0.572 0.123 0.000 0.276 15. Deepwater fish 0.000 0.053 0.050 0.333 0.013 0.046 0.000 0.000 0.000 0.000 0.000 16. Jellyfish 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 17. Cephalopods 0.150 0.230 0.840 0.320 0.000 0.009 0.070 0.040 0.000 0.000 0.010 18. Benth.P.Feeders 0.009 0.165 0.000 0.167 0.240 0.269 0.000 0.000 0.031 0.000 0.173 19. Infauna 0.000 0.000 0.000 0.000 0.400 0.000 0.000 0.001 0.010 0.000 0.200 20. Epifauna 0.015 0.000 0.000 0.000 0.250 0.000 0.000 0.000 0.001 0.000 0.072 21. Large Zoops 0.336 0.046 0.000 0.000 0.000 0.230 0.000 0.186 0.451 0.330 0.064 22. Herb. Zoops. 0.299 0.000 0.000 0.000 0.010 0.000 0.000 0.000 0.356 0.670 0.014 23. Phytoplankton 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 24. Detritus 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000        Import 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000        Sum 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000Predator100Table A4.2.  The estimated proportion of prey (nos. 1-24) eaten by predators (nos.12-22) during the 1950s.Prey12 13 14 15 16 17 18 19 20 21 22  1. Baleen whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  2. Toothed whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  3. Sperm whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  4. Beaked whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  5. Walrus&Bearded 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  6. Seals 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  7. Steller sea lion 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  8. Pisc. birds 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  9. Adult pollock 2+ 0.025 0.000 0.000 0.020 0.000 0.000 0.000 0.000 0.000 0.000 0.000 10. Juv. pollock 0-1 0.084 0.003 0.000 0.014 0.002 0.000 0.000 0.000 0.000 0.000 0.000 11. O.Dem.Fish 0.009 0.003 0.000 0.003 0.000 0.000 0.000 0.000 0.000 0.000 0.000 12. Large Flatfish 0.000 0.001 0.000 0.018 0.000 0.000 0.000 0.000 0.000 0.000 0.000 13. Small Flatfish 0.010 0.010 0.000 0.100 0.000 0.000 0.000 0.000 0.000 0.000 0.000 14. Pelagics 0.703 0.021 0.000 0.356 0.008 0.210 0.000 0.000 0.000 0.000 0.000 15. Deepwater fish 0.002 0.000 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000 16. Jellyfish 0.000 0.000 0.000 0.000 0.008 0.000 0.000 0.000 0.000 0.000 0.000 17. Cephalopods 0.004 0.000 0.000 0.110 0.000 0.190 0.000 0.000 0.000 0.000 0.000 18. Benth.P.Feeders 0.047 0.054 0.000 0.146 0.128 0.000 0.000 0.000 0.140 0.000 0.000 19. Infauna 0.002 0.600 0.005 0.102 0.000 0.000 0.300 0.000 0.300 0.000 0.000 20. Epifauna 0.009 0.084 0.000 0.020 0.000 0.000 0.037 0.000 0.008 0.000 0.000 21. Large Zoops 0.105 0.191 0.905 0.109 0.139 0.600 0.344 0.000 0.044 0.003 0.000 22. Herb. Zoops. 0.000 0.033 0.090 0.000 0.715 0.000 0.017 0.000 0.008 0.271 0.000 23. Phytoplankton 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.676 1.000 24. Detritus 0.000 0.000 0.000 0.000 0.000 0.000 0.302 1.000 0.500 0.050 0.000        Import 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000        Sum 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000Predator101Table A4.3.  The estimated proportion of prey (nos. 1-24) eaten by predators (nos. 1-11) during the 1980s.Prey1 2 3 4 5 6 7 8 9 10 11  1. Baleen whales 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  2. Toothed whales 0.000 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  3. Sperm whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  4. Beaked whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  5. Walrus&Bearded 0.000 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  6. Seals 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  7. Steller sea lion 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  8. Pisc. birds 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  9. Adult pollock 2+ 0.064 0.081 0.012 0.041 0.020 0.037 0.063 0.000 0.000 0.000 0.229 10. Juv. pollock 0-1 0.065 0.081 0.013 0.042 0.020 0.115 0.185 0.599 0.130 0.000 0.048 11. O.Dem.Fish 0.000 0.053 0.050 0.000 0.020 0.046 0.031 0.106 0.007 0.000 0.035 12. Large Flatfish 0.000 0.053 0.000 0.000 0.020 0.046 0.003 0.000 0.001 0.000 0.000 13. Small Flatfish 0.000 0.053 0.000 0.000 0.020 0.046 0.003 0.000 0.000 0.000 0.128 14. Pelagics 0.129 0.162 0.025 0.084 0.040 0.154 0.634 0.051 0.014 0.000 0.030 15. Deepwater fish 0.000 0.053 0.050 0.333 0.020 0.046 0.000 0.000 0.000 0.000 0.000 16. Jellyfish 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 17. Cephalopods 0.143 0.246 0.849 0.333 0.000 0.007 0.080 0.053 0.000 0.000 0.014 18. Benth.P.Feeders 0.049 0.165 0.000 0.167 0.250 0.270 0.000 0.000 0.031 0.000 0.244 19. Infauna 0.000 0.000 0.000 0.000 0.339 0.000 0.000 0.001 0.016 0.000 0.134 20. Epifauna 0.054 0.000 0.000 0.000 0.250 0.000 0.000 0.000 0.001 0.000 0.072 21. Large Zoops 0.253 0.046 0.000 0.000 0.000 0.232 0.000 0.188 0.441 0.330 0.064 22. Herb. Zoops. 0.243 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.357 0.670 0.000 23. Phytoplankton 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 24. Detritus 0.000 0.001 0.001 0.000 0.001 0.001 0.001 0.002 0.002 0.000 0.002        Import 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000        Sum 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000Predator102Table A4.4.  The estimated proportion of prey (nos. 1-24) eaten by predators (nos.12-22) during the 1980s.Prey12 13 14 15 16 17 18 19 20 21 22  1. Baleen whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  2. Toothed whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  3. Sperm whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  4. Beaked whales 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  5. Walrus&Bearded 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  6. Seals 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  7. Steller Sealion 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  8. Pisc. birds 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000  9. Adult pollock2+ 0.225 0.000 0.000 0.196 0.000 0.000 0.000 0.000 0.000 0.000 0.000 10. Juv. pollock0-1 0.523 0.016 0.000 0.087 0.010 0.000 0.000 0.000 0.000 0.000 0.000 11. O.Dem.Fish 0.006 0.002 0.000 0.003 0.000 0.000 0.000 0.000 0.000 0.000 0.000 12. Large Flatfish 0.000 0.001 0.000 0.018 0.000 0.000 0.000 0.000 0.000 0.000 0.000 13. Small Flatfish 0.013 0.011 0.000 0.100 0.000 0.000 0.000 0.000 0.000 0.000 0.000 14. Pelagics 0.061 0.008 0.000 0.100 0.000 0.200 0.000 0.000 0.000 0.000 0.000 15. Deepwater fish 0.002 0.000 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000 16. Jellyfish 0.000 0.000 0.000 0.000 0.008 0.000 0.000 0.000 0.000 0.000 0.000 17. Cephalopods 0.004 0.000 0.000 0.123 0.000 0.200 0.000 0.000 0.000 0.000 0.000 18. Benth.P.Feeders 0.048 0.055 0.000 0.236 0.128 0.000 0.000 0.000 0.004 0.000 0.000 19. Infauna 0.002 0.623 0.005 0.002 0.000 0.000 0.317 0.000 0.308 0.000 0.000 20. Epifauna 0.009 0.084 0.000 0.020 0.000 0.000 0.037 0.000 0.008 0.000 0.000 21. Large Zoops 0.105 0.198 0.905 0.111 0.139 0.600 0.338 0.000 0.044 0.003 0.000 22. Herb. Zoops. 0.000 0.000 0.090 0.000 0.715 0.000 0.000 0.000 0.000 0.213 0.000 23. Phytoplankton 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.676 1.000 24. Detritus 0.002 0.002 0.000 0.002 0.000 0.000 0.308 1.000 0.636 0.108 0.000        Import 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000        Sum 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000Predator103Appendix 5 – Niche Overlap TablesTable A5.1.  Estimated predator niche overlaps for all 24 groups of species in the1950s.  See section Niche Overlaps for explanations.  Overlaps > 0.40 are in boldcharacters.Predator Overlap 1950sGroup Name 1 2 3 4 5 6 7 8 9 10 11 12   1. Baleen whales 1.00 - - - - - - - - - - -   2. Toothed whales 1.00 1.00 - - - - - - - - - -   3. Sperm whales - - - - - - - - - - - -   4. Beaked whales - - - - - - - - - - - -   5. Walrus&Bearded 1.00 1.00 - - 1.00 - - - - - - -   6. Seals 1.00 1.00 - - 1.00 1.00 - - - - - -   7. Steller sea lion 1.00 1.00 - - 1.00 1.00 1.00 - - - - -   8. Pisc. birds - - - - - - - - - - - -   9. Adult pollock 2+ 0.00 0.00 - - 0.00 0.00 0.00 - 1.00 - - - 10. Juv. pollock 0-1 0.00 0.00 - - 0.00 0.00 0.00 - 0.43 1.00 - - 11. O.Dem.Fish 0.01 0.01 - - 0.01 0.01 0.01 - 0.95 0.45 1.00 - 12. Large Flatfish 0.06 0.06 - - 0.06 0.06 0.06 - 0.06 0.25 0.12 1.00 13. Small Flatfish 0.00 0.00 - - 0.00 0.00 0.00 - 0.98 0.35 0.96 0.09 14. Pelagics 0.00 0.00 - - 0.00 0.00 0.00 - 0.53 0.48 0.54 0.12 15. Deepwater fish 0.04 0.04 - - 0.04 0.04 0.04 - 0.03 0.07 0.15 0.51 16. Jellyfish - - - - - - - - - - - - 17. Cephalopods 0.00 0.00 - - 0.00 0.00 0.00 - 0.04 0.04 0.05 0.02 18. Benth.P.Feeders 0.00 0.00 - - 0.00 0.00 0.00 - 0.45 0.26 0.47 0.13 19. Infauna - - - - - - - - 0.06 0.06 0.07 0.06 20. Epifauna - - - - - - - - 0.16 0.10 0.18 0.07 21. Large Zoops - - - - - - - - 0.01 0.05 0.02 0.02 22. Herb. Zoops. - - - - - - - - 0.00 0.01 0.00 0.00 23. Phytoplankton - - - - - - - - - - - -Predator Overlap 1950s continuedGroup Name 13 14 15 16 17 18 19 20 21 22 23   1. Baleen whales - - - - - - - - - - -   2. Toothed whales - - - - - - - - - - -   3. Sperm whales - - - - - - - - - - -   4. Beaked whales - - - - - - - - - - -   5. Walrus&Bearded - - - - - - - - - - -   6. Seals - - - - - - - - - - -   7. Steller sea lion - - - - - - - - - - -   8. Pisc. birds - - - - - - - - - - -   9. Adult pollock 2+ - - - - - - - - - - - 10. Juv. pollock 0-1 - - - - - - - - - - - 11. O.Dem.Fish - - - - - - - - - - - 12. Large Flatfish - - - - - - - - - - - 13. Small Flatfish 1.00 - - - - - - - - - - 14. Pelagics 0.48 1.00 - - - - - - - - - 15. Deepwater fish 0.02 0.05 1.00 - - - - - - - - 16. Jellyfish - - - 1.00 - - - - - - - 17. Cephalopods 0.03 0.74 0.18 - 1.00 - - - - - - 18. Benth.P.Feeders 0.46 0.28 0.04 0.00 0.02 1.00 - - - - - 19. Infauna 0.08 0.04 - - 0.00 0.23 1.00 - - - - 20. Epifauna 0.19 0.10 0.00 - 0.01 0.16 0.98 1.00 - - - 21. Large Zoops 0.01 0.16 0.00 - 0.17 0.03 0.60 0.59 1.00 - - 22. Herb. Zoops. 0.00 0.00 - - 0.00 0.00 0.01 0.01 0.06 1.00 - 23. Phytoplankton - - - - - - - - - - -104Table A5.2.  Estimated predator niche overlaps for all 24 groups of species in the1950s.  See section Niche Overlaps for explanations.  Overlaps > 0.40 are in boldcharacters.Predator Overlap 1980sGroup Name 1 2 3 4 5 6 7 8 9 10 11 12   1. Baleen whales 1.00 - - - - - - - - - - -   2. Toothed whales 1.00 1.00 - - - - - - - - - -   3. Sperm whales - - - - - - - - - - - -   4. Beaked whales - - - - - - - - - - - -   5. Walrus&Bearded 1.00 1.00 - - 1.00 - - - - - - -   6. Seals 1.00 1.00 - - 1.00 1.00 - - - - - -   7. Steller sea lion - - - - - - - - - - - -   8. Pisc. birds - - - - - - - - - - - -   9. Adult pollock 2+ 0.00 0.00 - - 0.00 0.00 - - 1.00 - - - 10. Juv. pollock0-1 - - - - - - - - 0.16 1.00 - - 11. O.Dem.Fish 0.01 0.01 - - 0.01 0.01 - - 0.46 0.83 1.00 - 12. Large Flatfish 0.05 0.05 - - 0.05 0.05 - - 0.03 0.71 0.71 1.00 13. Small Flatfish 0.01 0.01 - - 0.01 0.01 - - 0.87 0.07 0.50 0.11 14. Pelagics 0.00 0.00 - - 0.00 0.00 - - 0.06 0.14 0.13 0.12 15. Deepwater fish 0.07 0.07 - - 0.07 0.07 - - 0.08 0.04 0.12 0.40 16. Jellyfish - - - - - - - - - - - - 17. Cephalopods 0.00 0.00 - - 0.00 0.00 - - 0.03 0.00 0.02 0.00 18. Benth.P.Feeders 0.00 0.00 - - 0.00 0.00 - - 0.56 0.64 0.90 0.68 19. Infauna - - - - - - - - 0.04 0.08 0.13 0.24 20. Epifauna - - - - - - - - 0.20 0.07 0.20 0.26 21. Large Zoops - - - - - - - - 0.01 0.43 0.42 0.40 22. Herb. Zoops. - - - - - - - - - 0.12 0.10 0.09 23. Phytoplankton - - - - - - - - - - - -Predator Overlap 1980s  continuedGroup Name 13 14 15 16 17 18 19 20 21 22 23   1. Baleen whales - - - - - - - - - - -   2. Toothed whales - - - - - - - - - - -   3. Sperm whales - - - - - - - - - - -   4. Beaked whales - - - - - - - - - - -   5. Walrus&Bearded - - - - - - - - - - -   6. Seals - - - - - - - - - - -   7. Steller Sealion - - - - - - - - - - -   8. Pisc. birds - - - - - - - - - - -   9. Adult pollock2+ - - - - - - - - - - - 10. Juv. pollock0-1 - - - - - - - - - - - 11. O.Dem.Fish - - - - - - - - - - - 12. Large Flatfish - - - - - - - - - - - 13. Small Flatfish 1.00 - - - - - - - - - - 14. Pelagics 0.04 1.00 - - - - - - - - - 15. Deepwater fish 0.04 0.02 1.00 - - - - - - - - 16. Jellyfish - - - 1.00 - - - - - - - 17. Cephalopods 0.02 0.98 0.07 - 1.00 - - - - - - 18. Benth.P.Feeders 0.66 0.12 0.08 0.00 0.01 1.00 - - - - - 19. Infauna 0.22 0.03 0.00 - - 0.37 1.00 - - - - 20. Epifauna 0.40 0.04 0.02 - 0.01 0.50 0.92 1.00 - - - 21. Large Zoops 0.03 0.37 0.00 - 0.29 0.36 0.23 0.21 1.00 - - 22. Herb. Zoops. - 0.02 - - - 0.07 0.01 0.00 0.14 1.00 - 23. Phytoplankton - - - - - - - - - - -105Table A5.3.  Estimated prey niche overlaps for all 24 groups of species in the 1950s.See section Niche Overlaps for explanations.  Overlaps > 0.40 are in boldcharacters.Prey Overlap 1950sGroup Name 1 2 3 4 5 6 7 8 9 10 11 12   1. Baleen whales 1.00 - - - - - - - - - - -   2. Toothed whales 0.48 1.00 - - - - - - - - - -   3. Sperm whales 0.28 0.48 1.00 - - - - - - - - -   4. Beaked whales 0.31 0.75 0.60 1.00 - - - - - - - -   5. Walrus&Bearded0.06 0.24 0.01 0.19 1.00 - - - - - - -   6. Seals 0.56 0.77 0.06 0.46 0.32 1.00 - - - - - -   7. Steller sea lion 0.33 0.62 0.15 0.33 0.07 0.52 1.00 - - - - -   8. Pisc. birds 0.54 0.72 0.13 0.34 0.08 0.72 0.89 1.00 - - - -   9. Adult pollock 2+ 0.93 0.24 0.01 0.09 0.07 0.53 0.19 0.43 1.00 - - - 10. Juv. pollock 0-1 0.76 0.04 - - 0.02 0.20 - 0.13 0.86 1.00 - - 11. O.Dem.Fish 0.37 0.71 0.06 0.36 0.68 0.76 0.52 0.63 0.29 0.08 1.00 - 12. Large Flatfish 0.43 0.68 0.07 0.33 0.11 0.66 0.96 0.96 0.32 0.06 0.62 1.00 13. Small Flatfish 0.24 0.08 0.00 0.04 0.80 0.21 0.03 0.12 0.29 0.18 0.53 0.09 14. Pelagics 0.61 0.08 - - 0.01 0.40 - 0.28 0.75 0.52 0.12 0.14 15. Deepwater fish 0.53 0.90 0.25 0.52 0.41 0.84 0.67 0.80 0.37 0.10 0.91 0.77 16. Jellyfish 0.65 0.08 - 0.06 0.09 0.18 0.01 0.07 0.72 0.95 0.12 0.05 17. Cephalopods 0.77 0.44 0.30 0.27 0.03 0.61 0.33 0.58 0.75 0.40 0.32 0.44 18. Benth.P.Feeders 0.43 0.06 - - 0.44 0.31 - 0.19 0.51 0.29 0.36 0.09 19. Infauna - - - - - - - - - - - - 20. Epifauna 0.06 0.09 - 0.07 0.48 0.17 - 0.02 0.08 0.04 0.33 0.03 21. Large Zoops 0.21 - - - 0.01 0.00 - 0.00 0.22 0.33 0.01 - 22. Herb. Zoops. - - - - - - - - - - - - 23. Phytoplankton - - - - - - - - - - - -Prey Overlap 1950 continuedGroup Name 13 14 15 16 17 18 19 20 21 22 23   1. Baleen whales - - - - - - - - - - -   2. Toothed whales - - - - - - - - - - -   3. Sperm whales - - - - - - - - - - -   4. Beaked whales - - - - - - - - - - -   5. Walrus&Bearded - - - - - - - - - - -   6. Seals - - - - - - - - - - -   7. Steller sea lion - - - - - - - - - - -   8. Pisc. birds - - - - - - - - - - -   9. Adult pollock 2+ - - - - - - - - - - - 10. Juv. pollock 0-1 - - - - - - - - - - - 11. O.Dem.Fish - - - - - - - - - - - 12. Large Flatfish - - - - - - - - - - - 13. Small Flatfish 1.00 - - - - - - - - - - 14. Pelagics 0.29 1.00 - - - - - - - - - 15. Deepwater fish 0.33 0.19 1.00 - - - - - - - - 16. Jellyfish 0.12 0.28 0.10 1.00 - - - - - - - 17. Cephalopods 0.28 0.86 0.51 0.17 1.00 - - - - - - 18. Benth.P.Feeders 0.70 0.56 0.28 0.14 0.56 1.00 - - - - - 19. Infauna - - - - - 0.46 1.00 - - - - 20. Epifauna 0.51 0.07 0.20 0.07 0.07 0.77 0.73 1.00 - - - 21. Large Zoops 0.02 0.04 - 0.36 0.00 0.05 0.07 0.06 1.00 - - 22. Herb. Zoops. - - - - - - - - 0.88 1.00 - 23. Phytoplankton - - - - - - - - - - -106Table A5.4.  Estimated prey niche overlaps for all 24 groups of species in the 1950s.See section Niche Overlaps for explanations.  Overlaps > 0.40 are in boldcharacters.Prey Overlap 1980sGroup Name 1 2 3 4 5 6 7 8 9 10 11 12   1. Baleen whales 1.00 - - - - - - - - - - -   2. Toothed whales 0.55 1.00 - - - - - - - - - -   3. Sperm whales 0.28 0.51 1.00 - - - - - - - - -   4. Beaked whales 0.33 0.74 0.61 1.00 - - - - - - - -   5. Walrus&Bearded 0.16 0.29 0.01 0.21 1.00 - - - - - - -   6. Seals 0.59 0.66 0.04 0.38 0.39 1.00 - - - - - -   7. Steller sea lion 0.35 0.49 0.15 0.26 0.09 0.40 1.00 - - - - -   8. Pisc. birds 0.34 0.31 0.10 0.14 0.05 0.43 0.35 1.00 - - - -   9. Adult pollock 2+ 0.82 0.16 0.00 0.04 0.06 0.50 0.08 0.43 1.00 - - - 10. Juv. pollock0-1 0.67 0.04 - - - 0.21 - 0.13 0.86 1.00 - - 11. O.Dem.Fish 0.33 0.55 0.04 0.28 0.67 0.64 0.15 0.16 0.18 0.06 1.00 - 12. Large Flatfish 0.33 0.35 0.03 0.15 0.11 0.45 0.38 0.90 0.34 0.08 0.40 1.00 13. Small Flatfish 0.19 0.07 - 0.03 0.72 0.21 0.02 0.11 0.26 0.13 0.39 0.09 14. Pelagics 0.50 0.09 - - 0.00 0.42 - 0.27 0.74 0.52 0.12 0.16 15. Deepwater fish 0.55 0.83 0.25 0.49 0.39 0.80 0.34 0.30 0.28 0.10 0.86 0.49 16. Jellyfish 0.60 0.08 - 0.05 0.08 0.19 0.00 0.07 0.73 0.95 0.11 0.06 17. Cephalopods 0.67 0.38 0.30 0.24 0.02 0.56 0.32 0.31 0.69 0.40 0.16 0.19 18. Benth.P.Feeders 0.36 0.07 - - 0.42 0.32 - 0.18 0.48 0.26 0.29 0.11 19. Infauna - 0.00 0.00 - 0.00 0.00 0.00 0.00 0.00 - 0.00 0.00 20. Epifauna 0.03 0.01 0.00 0.00 0.29 0.04 0.00 0.02 0.06 0.03 0.14 0.02 21. Large Zoops 0.15 - - - - 0.00 - 0.00 0.18 0.27 0.00 0.00 22. Herb. Zoops. - - - - - - - - - - - - 23. Phytoplankton - - - - - - - - - - - -Prey Overlap 1980s  continuedGroup Name 13 14 15 16 17 18 19 20 21 22 23   1. Baleen whales - - - - - - - - - - -   2. Toothed whales - - - - - - - - - - -   3. Sperm whales - - - - - - - - - - -   4. Beaked whales - - - - - - - - - - -   5. Walrus&Bearded - - - - - - - - - - -   6. Seals - - - - - - - - - - -   7. Steller sea lion - - - - - - - - - - -   8. Pisc. birds - - - - - - - - - - -   9. Adult pollock 2+ - - - - - - - - - - - 10. Juv. pollock 0-1 - - - - - - - - - - - 11. O.Dem.Fish - - - - - - - - - - - 12. Large Flatfish - - - - - - - - - - - 13. Small Flatfish 1.00 - - - - - - - - - - 14. Pelagics 0.29 1.00 - - - - - - - - - 15. Deepwater fish 0.14 0.21 1.00 - - - - - - - - 16. Jellyfish 0.07 0.28 0.13 1.00 - - - - - - - 17. Cephalopods 0.27 0.86 0.38 0.17 1.00 - - - - - - 18. Benth.P.Feeders 0.72 0.54 0.17 0.11 0.54 1.00 - - - - - 19. Infauna 0.00 - 0.00 - - 0.47 1.00 - - - - 20. Epifauna 0.43 0.06 0.02 0.01 0.06 0.76 0.85 1.00 - - - 21. Large Zoops 0.00 0.03 0.00 0.29 0.00 0.08 0.14 0.14 1.00 - - 22. Herb. Zoops. - - - - - - - - 0.89 1.00 - 23. Phytoplankton - - - - - - - - - - -


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items